id sid tid token lemma pos bracis-33639 1 1 modeling modeling NOUN bracis-33639 1 2 and and CCONJ bracis-33639 1 3     SPACE bracis-33639 1 4 predicting predict VERB bracis-33639 1 5 crimes crime NOUN bracis-33639 1 6 in in ADP bracis-33639 1 7     SPACE bracis-33639 1 8 the the DET bracis-33639 1 9     SPACE bracis-33639 1 10 city city NOUN bracis-33639 1 11 of of ADP bracis-33639 1 12     SPACE bracis-33639 1 13 são são PROPN bracis-33639 1 14 paulo paulo NOUN bracis-33639 1 15 using use VERB bracis-33639 1 16 graph graph NOUN bracis-33639 1 17 neural neural ADJ bracis-33639 1 18 networks network NOUN bracis-33639 1 19 | | ADV bracis-33639 1 20 springer springer NOUN bracis-33639 1 21 nature nature PROPN bracis-33639 1 22 link link PROPN bracis-33639 1 23 ( ( PUNCT bracis-33639 1 24 formerly formerly ADV bracis-33639 1 25 springerlink springerlink NOUN bracis-33639 1 26 ) ) PUNCT bracis-33639 1 27 skip skip VERB bracis-33639 1 28 to to ADP bracis-33639 1 29 main main ADJ bracis-33639 1 30 content content NOUN bracis-33639 1 31 advertisement advertisement NOUN bracis-33639 1 32 log log NOUN bracis-33639 1 33 in in ADP bracis-33639 1 34 menu menu NOUN bracis-33639 1 35 find find VERB bracis-33639 1 36 a a DET bracis-33639 1 37 journal journal NOUN bracis-33639 1 38 publish publish VERB bracis-33639 1 39 with with ADP bracis-33639 1 40 us we PRON bracis-33639 1 41 track track VERB bracis-33639 1 42 your your PRON bracis-33639 1 43 research research NOUN bracis-33639 1 44 search search NOUN bracis-33639 1 45 cart cart NOUN bracis-33639 1 46 home home NOUN bracis-33639 1 47 intelligent intelligent ADJ bracis-33639 1 48 systems system NOUN bracis-33639 1 49 conference conference NOUN bracis-33639 1 50 paper paper NOUN bracis-33639 1 51 modeling modeling NOUN bracis-33639 1 52 and and CCONJ bracis-33639 1 53     SPACE bracis-33639 1 54 predicting predict VERB bracis-33639 1 55 crimes crime NOUN bracis-33639 1 56 in in ADP bracis-33639 1 57     SPACE bracis-33639 1 58 the the DET bracis-33639 1 59     SPACE bracis-33639 1 60 city city NOUN bracis-33639 1 61 of of ADP bracis-33639 1 62     SPACE bracis-33639 1 63 são são PROPN bracis-33639 1 64 paulo paulo NOUN bracis-33639 1 65 using use VERB bracis-33639 1 66 graph graph NOUN bracis-33639 1 67 neural neural ADJ bracis-33639 1 68 networks network NOUN bracis-33639 1 69 conference conference NOUN bracis-33639 1 70 paper paper NOUN bracis-33639 1 71 first first ADV bracis-33639 1 72 online online ADV bracis-33639 1 73 : : PUNCT bracis-33639 1 74 30 30 NUM bracis-33639 1 75 january january NOUN bracis-33639 1 76 2025 2025 NUM bracis-33639 1 77 pp pp CCONJ bracis-33639 1 78 372–386 372–386 NUM bracis-33639 1 79 cite cite VERB bracis-33639 1 80 this this DET bracis-33639 1 81 conference conference NOUN bracis-33639 1 82 paper paper NOUN bracis-33639 1 83 access access NOUN bracis-33639 1 84 provided provide VERB bracis-33639 1 85 by by ADP bracis-33639 1 86 university university PROPN bracis-33639 1 87 of of ADP bracis-33639 1 88 notre notre PROPN bracis-33639 1 89 dame dame PROPN bracis-33639 1 90 hesburgh hesburgh PROPN bracis-33639 1 91 library library PROPN bracis-33639 1 92 download download PROPN bracis-33639 1 93 book book NOUN bracis-33639 1 94 pdf pdf PROPN bracis-33639 1 95 download download NOUN bracis-33639 1 96 book book NOUN bracis-33639 1 97 epub epub PROPN bracis-33639 1 98 intelligent intelligent ADJ bracis-33639 1 99 systems system NOUN bracis-33639 1 100 ( ( PUNCT bracis-33639 1 101 bracis bracis NOUN bracis-33639 1 102 2024 2024 NUM bracis-33639 1 103 ) ) PUNCT bracis-33639 1 104 modeling modeling NOUN bracis-33639 1 105 and and CCONJ bracis-33639 1 106     SPACE bracis-33639 1 107 predicting predict VERB bracis-33639 1 108 crimes crime NOUN bracis-33639 1 109 in in ADP bracis-33639 1 110     SPACE bracis-33639 1 111 the the DET bracis-33639 1 112     SPACE bracis-33639 1 113 city city NOUN bracis-33639 1 114 of of ADP bracis-33639 1 115     SPACE bracis-33639 1 116 são são PROPN bracis-33639 1 117 paulo paulo NOUN bracis-33639 1 118 using use VERB bracis-33639 1 119 graph graph NOUN bracis-33639 1 120 neural neural ADJ bracis-33639 1 121 networks network NOUN bracis-33639 1 122 download download NOUN bracis-33639 1 123 book book NOUN bracis-33639 1 124 pdf pdf PROPN bracis-33639 1 125 download download NOUN bracis-33639 1 126 book book PROPN bracis-33639 1 127 epub epub PROPN bracis-33639 1 128 waqar waqar PROPN bracis-33639 1 129 hassan hassan PROPN bracis-33639 1 130     SPACE bracis-33639 1 131 orcid orcid NOUN bracis-33639 1 132 : : PUNCT bracis-33639 2 1 orcid.org/0000-0002-6956-82499 orcid.org/0000-0002-6956-82499 PROPN bracis-33639 2 2 , , PUNCT bracis-33639 2 3 marvin marvin PROPN bracis-33639 2 4 mendes mendes PROPN bracis-33639 2 5 cabral cabral PROPN bracis-33639 2 6     SPACE bracis-33639 2 7 orcid orcid NOUN bracis-33639 2 8 : : PUNCT bracis-33639 2 9 orcid.org/0009-0006-5144-03909 orcid.org/0009-0006-5144-03909 ADJ bracis-33639 2 10 , , PUNCT bracis-33639 2 11 thiago thiago PROPN bracis-33639 2 12 rodrigo rodrigo PROPN bracis-33639 2 13 ramos ramos PROPN bracis-33639 2 14     SPACE bracis-33639 2 15 orcid orcid PROPN bracis-33639 2 16 : : PUNCT bracis-33639 2 17 orcid.org/0000-0001-6271-709910 orcid.org/0000-0001-6271-709910 NOUN bracis-33639 2 18 , , PUNCT bracis-33639 2 19 antonio antonio PROPN bracis-33639 2 20 castelo castelo PROPN bracis-33639 2 21 filho filho PROPN bracis-33639 2 22     SPACE bracis-33639 2 23 orcid orcid NOUN bracis-33639 2 24 : : PUNCT bracis-33639 2 25 orcid.org/0000-0001-8009-45779 orcid.org/0000-0001-8009-45779 PROPN bracis-33639 2 26 & & CCONJ bracis-33639 2 27 … … PUNCT bracis-33639 2 28 luis luis PROPN bracis-33639 2 29 gustavo gustavo PROPN bracis-33639 2 30 nonato nonato PROPN bracis-33639 2 31     SPACE bracis-33639 2 32 orcid orcid NOUN bracis-33639 2 33 : : PUNCT bracis-33639 2 34 orcid.org/0000-0002-8514-80339 orcid.org/0000-0002-8514-80339 NOUN bracis-33639 2 35     SPACE bracis-33639 2 36 show show VERB bracis-33639 2 37 authors author NOUN bracis-33639 2 38 part part NOUN bracis-33639 2 39 of of ADP bracis-33639 2 40 the the DET bracis-33639 2 41 book book NOUN bracis-33639 2 42 series series NOUN bracis-33639 2 43 : : PUNCT bracis-33639 2 44 lecture lecture NOUN bracis-33639 2 45 notes note NOUN bracis-33639 2 46 in in ADP bracis-33639 2 47 computer computer NOUN bracis-33639 2 48 science science NOUN bracis-33639 2 49 ( ( PUNCT bracis-33639 2 50 ( ( PUNCT bracis-33639 2 51 lnai lnai ADJ bracis-33639 2 52 , , PUNCT bracis-33639 2 53 volume volume NOUN bracis-33639 2 54 15414 15414 NUM bracis-33639 2 55 ) ) PUNCT bracis-33639 2 56 ) ) PUNCT bracis-33639 2 57 included include VERB bracis-33639 2 58 in in ADP bracis-33639 2 59 the the DET bracis-33639 2 60 following follow VERB bracis-33639 2 61 conference conference NOUN bracis-33639 2 62 series series NOUN bracis-33639 2 63 : : PUNCT bracis-33639 2 64 brazilian brazilian ADJ bracis-33639 2 65 conference conference NOUN bracis-33639 2 66 on on ADP bracis-33639 2 67 intelligent intelligent ADJ bracis-33639 2 68 systems system NOUN bracis-33639 2 69 473 473 NUM bracis-33639 2 70 accesses access VERB bracis-33639 2 71 2 2 NUM bracis-33639 2 72 citations citation NOUN bracis-33639 2 73 abstract abstract ADJ bracis-33639 2 74 crime crime NOUN bracis-33639 2 75 prediction prediction NOUN bracis-33639 2 76 is be AUX bracis-33639 2 77 a a DET bracis-33639 2 78 critical critical ADJ bracis-33639 2 79 research research NOUN bracis-33639 2 80 area area NOUN bracis-33639 2 81 for for ADP bracis-33639 2 82 enhancing enhance VERB bracis-33639 2 83 public public ADJ bracis-33639 2 84 safety safety NOUN bracis-33639 2 85 and and CCONJ bracis-33639 2 86 optimizing optimize VERB bracis-33639 2 87 law law NOUN bracis-33639 2 88 enforcement enforcement NOUN bracis-33639 2 89 resource resource NOUN bracis-33639 2 90 allocation allocation NOUN bracis-33639 2 91 , , PUNCT bracis-33639 2 92 and and CCONJ bracis-33639 2 93 machine machine NOUN bracis-33639 2 94 learning learning NOUN bracis-33639 2 95 techniques technique NOUN bracis-33639 2 96 have have AUX bracis-33639 2 97 had have VERB bracis-33639 2 98 a a DET bracis-33639 2 99 significant significant ADJ bracis-33639 2 100 impact impact NOUN bracis-33639 2 101 in in ADP bracis-33639 2 102 this this DET bracis-33639 2 103 field field NOUN bracis-33639 2 104 . . PUNCT bracis-33639 3 1 traditional traditional ADJ bracis-33639 3 2 machine machine NOUN bracis-33639 3 3 learning learning NOUN bracis-33639 3 4 models model NOUN bracis-33639 3 5 have have AUX bracis-33639 3 6 long long ADV bracis-33639 3 7 struggled struggle VERB bracis-33639 3 8 to to PART bracis-33639 3 9 capture capture VERB bracis-33639 3 10 complex complex ADJ bracis-33639 3 11 crime crime NOUN bracis-33639 3 12 patterns pattern NOUN bracis-33639 3 13 , , PUNCT bracis-33639 3 14 primarily primarily ADV bracis-33639 3 15 due due ADP bracis-33639 3 16 to to ADP bracis-33639 3 17 the the DET bracis-33639 3 18 intricate intricate ADJ bracis-33639 3 19 interdependence interdependence NOUN bracis-33639 3 20 of of ADP bracis-33639 3 21 spatial spatial ADJ bracis-33639 3 22 and and CCONJ bracis-33639 3 23 temporal temporal ADJ bracis-33639 3 24 data datum NOUN bracis-33639 3 25 . . PUNCT bracis-33639 4 1 however however ADV bracis-33639 4 2 , , PUNCT bracis-33639 4 3 recent recent ADJ bracis-33639 4 4 advancements advancement NOUN bracis-33639 4 5 in in ADP bracis-33639 4 6 machine machine NOUN bracis-33639 4 7 learning learning NOUN bracis-33639 4 8 , , PUNCT bracis-33639 4 9 particularly particularly ADV bracis-33639 4 10 with with ADP bracis-33639 4 11 graph graph NOUN bracis-33639 4 12 neural neural ADJ bracis-33639 4 13 networks network NOUN bracis-33639 4 14 ( ( PUNCT bracis-33639 4 15 gnns gnns NOUN bracis-33639 4 16 ) ) PUNCT bracis-33639 4 17 , , PUNCT bracis-33639 4 18 offer offer VERB bracis-33639 4 19 a a DET bracis-33639 4 20 new new ADJ bracis-33639 4 21 perspective perspective NOUN bracis-33639 4 22 . . PUNCT bracis-33639 5 1 gnns gnns PROPN bracis-33639 5 2 have have AUX bracis-33639 5 3 demonstrated demonstrate VERB bracis-33639 5 4 remarkable remarkable ADJ bracis-33639 5 5 success success NOUN bracis-33639 5 6 in in ADP bracis-33639 5 7 various various ADJ bracis-33639 5 8 applications application NOUN bracis-33639 5 9 and and CCONJ bracis-33639 5 10 they they PRON bracis-33639 5 11 can can AUX bracis-33639 5 12 also also ADV bracis-33639 5 13 play play VERB bracis-33639 5 14 a a DET bracis-33639 5 15 significant significant ADJ bracis-33639 5 16 role role NOUN bracis-33639 5 17 in in ADP bracis-33639 5 18 crime crime NOUN bracis-33639 5 19 analysis analysis NOUN bracis-33639 5 20 and and CCONJ bracis-33639 5 21 prediction prediction NOUN bracis-33639 5 22 . . PUNCT bracis-33639 6 1 therefore therefore ADV bracis-33639 6 2 , , PUNCT bracis-33639 6 3 in in ADP bracis-33639 6 4 this this DET bracis-33639 6 5 work work NOUN bracis-33639 6 6 , , PUNCT bracis-33639 6 7 we we PRON bracis-33639 6 8 explore explore VERB bracis-33639 6 9 such such DET bracis-33639 6 10 a a DET bracis-33639 6 11 potential potential NOUN bracis-33639 6 12 by by ADP bracis-33639 6 13 examining examine VERB bracis-33639 6 14 two two NUM bracis-33639 6 15 distinct distinct ADJ bracis-33639 6 16 spatiotemporal spatiotemporal ADJ bracis-33639 6 17 gnn gnn PROPN bracis-33639 6 18 architectures architecture NOUN bracis-33639 6 19 , , PUNCT bracis-33639 6 20 namely namely ADV bracis-33639 6 21 dynamic dynamic ADJ bracis-33639 6 22 self self NOUN bracis-33639 6 23 - - PUNCT bracis-33639 6 24 attention attention NOUN bracis-33639 6 25 network network NOUN bracis-33639 6 26 ( ( PUNCT bracis-33639 6 27 dysat dysat PROPN bracis-33639 6 28 ) ) PUNCT bracis-33639 6 29 and and CCONJ bracis-33639 6 30 evolving evolve VERB bracis-33639 6 31 graph graph NOUN bracis-33639 6 32 convolutional convolutional ADJ bracis-33639 6 33 network network NOUN bracis-33639 6 34 ( ( PUNCT bracis-33639 6 35 evolvegcn evolvegcn PROPN bracis-33639 6 36 ) ) PUNCT bracis-33639 6 37 , , PUNCT bracis-33639 6 38 assessing assess VERB bracis-33639 6 39 and and CCONJ bracis-33639 6 40 comparing compare VERB bracis-33639 6 41 their their PRON bracis-33639 6 42 effectiveness effectiveness NOUN bracis-33639 6 43 for for ADP bracis-33639 6 44 crime crime NOUN bracis-33639 6 45 prediction prediction NOUN bracis-33639 6 46 . . PUNCT bracis-33639 7 1 moreover moreover ADV bracis-33639 7 2 , , PUNCT bracis-33639 7 3 we we PRON bracis-33639 7 4 propose propose VERB bracis-33639 7 5 a a DET bracis-33639 7 6 data datum NOUN bracis-33639 7 7 modeling model VERB bracis-33639 7 8 framework framework NOUN bracis-33639 7 9 that that PRON bracis-33639 7 10 integrates integrate VERB bracis-33639 7 11 crime crime NOUN bracis-33639 7 12 , , PUNCT bracis-33639 7 13 street street NOUN bracis-33639 7 14 map map NOUN bracis-33639 7 15 graphs graph NOUN bracis-33639 7 16 , , PUNCT bracis-33639 7 17 and and CCONJ bracis-33639 7 18 urban urban ADJ bracis-33639 7 19 data datum NOUN bracis-33639 7 20 , , PUNCT bracis-33639 7 21 which which PRON bracis-33639 7 22 is be AUX bracis-33639 7 23 fundamental fundamental ADJ bracis-33639 7 24 to to PART bracis-33639 7 25 properly properly ADV bracis-33639 7 26 train train VERB bracis-33639 7 27 the the DET bracis-33639 7 28 gnn gnn PROPN bracis-33639 7 29 models model NOUN bracis-33639 7 30 . . PUNCT bracis-33639 8 1 as as ADV bracis-33639 8 2 far far ADV bracis-33639 8 3 as as SCONJ bracis-33639 8 4 we we PRON bracis-33639 8 5 know know VERB bracis-33639 8 6 , , PUNCT bracis-33639 8 7 there there PRON bracis-33639 8 8 is be VERB bracis-33639 8 9 no no DET bracis-33639 8 10 consolidated consolidated ADJ bracis-33639 8 11 methodology methodology NOUN bracis-33639 8 12 to to PART bracis-33639 8 13 integrate integrate VERB bracis-33639 8 14 those those DET bracis-33639 8 15 three three NUM bracis-33639 8 16 modalities modality NOUN bracis-33639 8 17 of of ADP bracis-33639 8 18 data datum NOUN bracis-33639 8 19 , , PUNCT bracis-33639 8 20 being be AUX bracis-33639 8 21 a a DET bracis-33639 8 22 relevant relevant ADJ bracis-33639 8 23 contribution contribution NOUN bracis-33639 8 24 of of ADP bracis-33639 8 25 this this DET bracis-33639 8 26 work work NOUN bracis-33639 8 27 . . PUNCT bracis-33639 9 1 our our PRON bracis-33639 9 2 findings finding NOUN bracis-33639 9 3 underscore underscore VERB bracis-33639 9 4 the the DET bracis-33639 9 5 effectiveness effectiveness NOUN bracis-33639 9 6 of of ADP bracis-33639 9 7 gnns gnns NOUN bracis-33639 9 8 in in ADP bracis-33639 9 9 crime crime NOUN bracis-33639 9 10 prediction prediction NOUN bracis-33639 9 11 tasks task NOUN bracis-33639 9 12 , , PUNCT bracis-33639 9 13 offering offer VERB bracis-33639 9 14 valuable valuable ADJ bracis-33639 9 15 insights insight NOUN bracis-33639 9 16 for for ADP bracis-33639 9 17 researchers researcher NOUN bracis-33639 9 18 and and CCONJ bracis-33639 9 19 practitioners practitioner NOUN bracis-33639 9 20 in in ADP bracis-33639 9 21 the the DET bracis-33639 9 22 field field NOUN bracis-33639 9 23 of of ADP bracis-33639 9 24 crime crime NOUN bracis-33639 9 25 prevention prevention NOUN bracis-33639 9 26 and and CCONJ bracis-33639 9 27 public public ADJ bracis-33639 9 28 safety safety NOUN bracis-33639 9 29 enhancement enhancement NOUN bracis-33639 9 30 . . PUNCT bracis-33639 10 1 access access NOUN bracis-33639 10 2 provided provide VERB bracis-33639 10 3 by by ADP bracis-33639 10 4 university university PROPN bracis-33639 10 5 of of ADP bracis-33639 10 6 notre notre PROPN bracis-33639 10 7 dame dame PROPN bracis-33639 10 8 hesburgh hesburgh PROPN bracis-33639 10 9 library library PROPN bracis-33639 10 10 . . PUNCT bracis-33639 11 1 download download PROPN bracis-33639 11 2 conference conference NOUN bracis-33639 11 3 paper paper NOUN bracis-33639 11 4 pdf pdf NOUN bracis-33639 11 5 similar similar ADJ bracis-33639 11 6 content content NOUN bracis-33639 11 7 being be AUX bracis-33639 11 8 viewed view VERB bracis-33639 11 9 by by ADP bracis-33639 11 10 others other NOUN bracis-33639 11 11 crime crime NOUN bracis-33639 11 12 prediction prediction NOUN bracis-33639 11 13 with with ADP bracis-33639 11 14 graph graph NOUN bracis-33639 11 15 neural neural ADJ bracis-33639 11 16 networks network NOUN bracis-33639 11 17 and and CCONJ bracis-33639 11 18 multivariate multivariate VERB bracis-33639 11 19 normal normal ADJ bracis-33639 11 20 distributions distribution NOUN bracis-33639 11 21 article article NOUN bracis-33639 11 22 01 01 NUM bracis-33639 11 23 august august PROPN bracis-33639 11 24 2022 2022 NUM bracis-33639 11 25 ai ai ADJ bracis-33639 11 26 - - PUNCT bracis-33639 11 27 driven drive VERB bracis-33639 11 28 crime crime NOUN bracis-33639 11 29 prediction prediction NOUN bracis-33639 11 30 : : PUNCT bracis-33639 11 31 a a DET bracis-33639 11 32 systematic systematic ADJ bracis-33639 11 33 literature literature NOUN bracis-33639 11 34 review review NOUN bracis-33639 11 35 article article NOUN bracis-33639 11 36 16 16 NUM bracis-33639 11 37 april april PROPN bracis-33639 11 38 2025 2025 NUM bracis-33639 11 39 deep deep ADJ bracis-33639 11 40 temporal temporal ADJ bracis-33639 11 41 multi multi ADJ bracis-33639 11 42 - - ADJ bracis-33639 11 43 graph graph ADJ bracis-33639 11 44 convolutional convolutional ADJ bracis-33639 11 45 network network NOUN bracis-33639 11 46 for for ADP bracis-33639 11 47 crime crime NOUN bracis-33639 11 48 prediction prediction NOUN bracis-33639 11 49 chapter chapter NOUN bracis-33639 11 50 © © NOUN bracis-33639 11 51 2020 2020 NUM bracis-33639 11 52 1 1 NUM bracis-33639 11 53 introduction introduction NOUN bracis-33639 11 54 crime crime NOUN bracis-33639 11 55 prediction prediction NOUN bracis-33639 11 56 has have AUX bracis-33639 11 57 long long ADV bracis-33639 11 58 been be AUX bracis-33639 11 59 a a DET bracis-33639 11 60 topic topic NOUN bracis-33639 11 61 of of ADP bracis-33639 11 62 great great ADJ bracis-33639 11 63 interest interest NOUN bracis-33639 11 64 , , PUNCT bracis-33639 11 65 supporting support VERB bracis-33639 11 66 law law NOUN bracis-33639 11 67 enforcement enforcement NOUN bracis-33639 11 68 agencies agency NOUN bracis-33639 11 69 in in ADP bracis-33639 11 70 their their PRON bracis-33639 11 71 decision decision NOUN bracis-33639 11 72 - - PUNCT bracis-33639 11 73 making make VERB bracis-33639 11 74 processes process NOUN bracis-33639 11 75     SPACE bracis-33639 12 1 [ [ X bracis-33639 12 2 11 11 NUM bracis-33639 12 3 ] ] PUNCT bracis-33639 12 4 . . PUNCT bracis-33639 13 1 in in ADP bracis-33639 13 2 this this DET bracis-33639 13 3 context context NOUN bracis-33639 13 4 , , PUNCT bracis-33639 13 5 machine machine NOUN bracis-33639 13 6 learning learning NOUN bracis-33639 13 7 ( ( PUNCT bracis-33639 13 8 ml ml NOUN bracis-33639 13 9 ) ) PUNCT bracis-33639 13 10 techniques technique NOUN bracis-33639 13 11 have have AUX bracis-33639 13 12 played play VERB bracis-33639 13 13 a a DET bracis-33639 13 14 major major ADJ bracis-33639 13 15 role role NOUN bracis-33639 13 16 , , PUNCT bracis-33639 13 17 providing provide VERB bracis-33639 13 18 a a DET bracis-33639 13 19 multitude multitude NOUN bracis-33639 13 20 of of ADP bracis-33639 13 21 analytical analytical ADJ bracis-33639 13 22 tools tool NOUN bracis-33639 13 23 to to PART bracis-33639 13 24 scrutinize scrutinize VERB bracis-33639 13 25 and and CCONJ bracis-33639 13 26 predict predict VERB bracis-33639 13 27 crime crime NOUN bracis-33639 13 28 - - PUNCT bracis-33639 13 29 related relate VERB bracis-33639 13 30 phenomena phenomenon NOUN bracis-33639 13 31     SPACE bracis-33639 14 1 [ [ X bracis-33639 14 2 1 1 NUM bracis-33639 14 3 , , PUNCT bracis-33639 14 4 8 8 NUM bracis-33639 14 5 , , PUNCT bracis-33639 14 6 17 17 NUM bracis-33639 14 7 , , PUNCT bracis-33639 14 8 20 20 NUM bracis-33639 14 9 ] ] PUNCT bracis-33639 14 10 . . PUNCT bracis-33639 15 1 however however ADV bracis-33639 15 2 , , PUNCT bracis-33639 15 3 traditional traditional ADJ bracis-33639 15 4 machine machine NOUN bracis-33639 15 5 learning learn VERB bracis-33639 15 6 algorithms algorithm NOUN bracis-33639 15 7 often often ADV bracis-33639 15 8 fail fail VERB bracis-33639 15 9 to to PART bracis-33639 15 10 capture capture VERB bracis-33639 15 11 complex complex ADJ bracis-33639 15 12 spatiotemporal spatiotemporal ADJ bracis-33639 15 13 patterns pattern NOUN bracis-33639 15 14 inherent inherent ADJ bracis-33639 15 15 in in ADP bracis-33639 15 16 non non ADJ bracis-33639 15 17 - - ADJ bracis-33639 15 18 random random ADJ bracis-33639 15 19 crime crime NOUN bracis-33639 15 20 events event NOUN bracis-33639 15 21     SPACE bracis-33639 16 1 [ [ X bracis-33639 16 2 13 13 NUM bracis-33639 16 3 ] ] PUNCT bracis-33639 16 4 . . PUNCT bracis-33639 17 1 a a DET bracis-33639 17 2 primary primary ADJ bracis-33639 17 3 reason reason NOUN bracis-33639 17 4 for for ADP bracis-33639 17 5 their their PRON bracis-33639 17 6 limited limited ADJ bracis-33639 17 7 performance performance NOUN bracis-33639 17 8 is be AUX bracis-33639 17 9 that that SCONJ bracis-33639 17 10 these these DET bracis-33639 17 11 models model NOUN bracis-33639 17 12 handle handle VERB bracis-33639 17 13 each each DET bracis-33639 17 14 data datum NOUN bracis-33639 17 15 point point NOUN bracis-33639 17 16 independently independently ADV bracis-33639 17 17 , , PUNCT bracis-33639 17 18 disregarding disregard VERB bracis-33639 17 19 the the DET bracis-33639 17 20 interconnected interconnected ADJ bracis-33639 17 21 nature nature NOUN bracis-33639 17 22 of of ADP bracis-33639 17 23 events event NOUN bracis-33639 17 24 over over ADP bracis-33639 17 25 time time NOUN bracis-33639 17 26 and and CCONJ bracis-33639 17 27 space space NOUN bracis-33639 17 28     SPACE bracis-33639 18 1 [ [ X bracis-33639 18 2 14 14 NUM bracis-33639 18 3 ] ] PUNCT bracis-33639 18 4 . . PUNCT bracis-33639 19 1 recent recent ADJ bracis-33639 19 2 advancements advancement NOUN bracis-33639 19 3 in in ADP bracis-33639 19 4 machine machine NOUN bracis-33639 19 5 learning learning NOUN bracis-33639 19 6 , , PUNCT bracis-33639 19 7 particularly particularly ADV bracis-33639 19 8 in in ADP bracis-33639 19 9 graph graph NOUN bracis-33639 19 10 neural neural ADJ bracis-33639 19 11 networks network NOUN bracis-33639 19 12 ( ( PUNCT bracis-33639 19 13 gnns gnns NOUN bracis-33639 19 14 ) ) PUNCT bracis-33639 20 1 [ [ X bracis-33639 20 2 31 31 NUM bracis-33639 20 3 ] ] PUNCT bracis-33639 20 4 , , PUNCT bracis-33639 20 5 bring bring VERB bracis-33639 20 6 new new ADJ bracis-33639 20 7 perspectives perspective NOUN bracis-33639 20 8 to to ADP bracis-33639 20 9 these these DET bracis-33639 20 10 challenging challenging ADJ bracis-33639 20 11 scenarios scenario NOUN bracis-33639 20 12 . . PUNCT bracis-33639 21 1 specifically specifically ADV bracis-33639 21 2 , , PUNCT bracis-33639 21 3 gnns gnn NOUN bracis-33639 21 4 have have AUX bracis-33639 21 5 demonstrated demonstrate VERB bracis-33639 21 6 remarkable remarkable ADJ bracis-33639 21 7 success success NOUN bracis-33639 21 8 across across ADP bracis-33639 21 9 various various ADJ bracis-33639 21 10 applications application NOUN bracis-33639 21 11 involving involve VERB bracis-33639 21 12 spatiotemporal spatiotemporal ADJ bracis-33639 21 13 data datum NOUN bracis-33639 21 14 , , PUNCT bracis-33639 21 15 including include VERB bracis-33639 21 16 social social ADJ bracis-33639 21 17 network network NOUN bracis-33639 21 18 evolution evolution NOUN bracis-33639 21 19     SPACE bracis-33639 22 1 [ [ X bracis-33639 22 2 7 7 NUM bracis-33639 22 3 ] ] PUNCT bracis-33639 22 4 , , PUNCT bracis-33639 22 5 traffic traffic NOUN bracis-33639 22 6 prediction prediction NOUN bracis-33639 22 7     SPACE bracis-33639 23 1 [ [ X bracis-33639 23 2 12 12 NUM bracis-33639 23 3 ] ] PUNCT bracis-33639 23 4 , , PUNCT bracis-33639 23 5 and and CCONJ bracis-33639 23 6 identification identification NOUN bracis-33639 23 7 of of ADP bracis-33639 23 8 urban urban ADJ bracis-33639 23 9 structural structural ADJ bracis-33639 23 10 patterns pattern NOUN bracis-33639 23 11     SPACE bracis-33639 24 1 [ [ X bracis-33639 24 2 30 30 NUM bracis-33639 24 3 ] ] PUNCT bracis-33639 24 4 . . PUNCT bracis-33639 25 1 despite despite SCONJ bracis-33639 25 2 the the DET bracis-33639 25 3 success success NOUN bracis-33639 25 4 in in ADP bracis-33639 25 5 various various ADJ bracis-33639 25 6 domains domain NOUN bracis-33639 25 7 , , PUNCT bracis-33639 25 8 graph graph NOUN bracis-33639 25 9 neural neural ADJ bracis-33639 25 10 networks network NOUN bracis-33639 25 11 have have AUX bracis-33639 25 12 not not PART bracis-33639 25 13 yet yet ADV bracis-33639 25 14 been be AUX bracis-33639 25 15 fully fully ADV bracis-33639 25 16 explored explore VERB bracis-33639 25 17 for for ADP bracis-33639 25 18 crime crime NOUN bracis-33639 25 19 prediction prediction NOUN bracis-33639 25 20 tasks task NOUN bracis-33639 25 21 , , PUNCT bracis-33639 25 22     SPACE bracis-33639 25 23 where where SCONJ bracis-33639 25 24 traditional traditional ADJ bracis-33639 25 25 ml ml NOUN bracis-33639 25 26 models model NOUN bracis-33639 25 27 still still ADV bracis-33639 25 28 dominate dominate VERB bracis-33639 25 29 the the DET bracis-33639 25 30 field field NOUN bracis-33639 25 31 . . PUNCT bracis-33639 26 1 among among ADP bracis-33639 26 2 the the DET bracis-33639 26 3 few few ADJ bracis-33639 26 4 works work NOUN bracis-33639 26 5 that that PRON bracis-33639 26 6 utilize utilize VERB bracis-33639 26 7 gnns gnns NOUN bracis-33639 26 8 for for ADP bracis-33639 26 9 crime crime NOUN bracis-33639 26 10 analysis analysis NOUN bracis-33639 26 11 and and CCONJ bracis-33639 26 12 forecasting forecasting NOUN bracis-33639 26 13 is be AUX bracis-33639 26 14 the the DET bracis-33639 26 15 model model NOUN bracis-33639 26 16 proposed propose VERB bracis-33639 26 17 by by ADP bracis-33639 26 18 han han PROPN bracis-33639 26 19 et et PROPN bracis-33639 26 20 al al PROPN bracis-33639 26 21 . . PROPN bracis-33639 26 22     SPACE bracis-33639 27 1 [ [ X bracis-33639 27 2 10 10 NUM bracis-33639 27 3 ] ] PUNCT bracis-33639 27 4 , , PUNCT bracis-33639 27 5 which which PRON bracis-33639 27 6 leverages leverage VERB bracis-33639 27 7 graph graph NOUN bracis-33639 27 8 convolutional convolutional ADJ bracis-33639 27 9 networks network NOUN bracis-33639 27 10 ( ( PUNCT bracis-33639 27 11 gcns gcns PROPN bracis-33639 27 12 ) ) PUNCT bracis-33639 27 13 and and CCONJ bracis-33639 27 14 long long ADJ bracis-33639 27 15 short short ADJ bracis-33639 27 16 - - PUNCT bracis-33639 27 17 term term NOUN bracis-33639 27 18 memory memory NOUN bracis-33639 27 19 ( ( PUNCT bracis-33639 27 20 lstm lstm NOUN bracis-33639 27 21 ) ) PUNCT bracis-33639 27 22 networks network NOUN bracis-33639 27 23 to to PART bracis-33639 27 24 capture capture VERB bracis-33639 27 25 spatial spatial ADJ bracis-33639 27 26 dependencies dependency NOUN bracis-33639 27 27 and and CCONJ bracis-33639 27 28 temporal temporal ADJ bracis-33639 27 29 patterns pattern NOUN bracis-33639 27 30 of of ADP bracis-33639 27 31 crimes crime NOUN bracis-33639 27 32 . . PUNCT bracis-33639 28 1 a a DET bracis-33639 28 2 similar similar ADJ bracis-33639 28 3 methodology methodology NOUN bracis-33639 28 4 has have AUX bracis-33639 28 5 been be AUX bracis-33639 28 6 proposed propose VERB bracis-33639 28 7 by by ADP bracis-33639 28 8 jin jin PROPN bracis-33639 28 9 et et PROPN bracis-33639 28 10 al al PROPN bracis-33639 28 11 . . PUNCT bracis-33639 28 12     SPACE bracis-33639 29 1 [ [ X bracis-33639 29 2 13 13 NUM bracis-33639 29 3 ] ] PUNCT bracis-33639 29 4 , , PUNCT bracis-33639 29 5 where where SCONJ bracis-33639 29 6 graph graph NOUN bracis-33639 29 7 convolutional convolutional ADJ bracis-33639 29 8 networks network NOUN bracis-33639 29 9 and and CCONJ bracis-33639 29 10 recurrent recurrent ADJ bracis-33639 29 11 neural neural ADJ bracis-33639 29 12 networks network NOUN bracis-33639 29 13 ( ( PUNCT bracis-33639 29 14 rnns rnns PROPN bracis-33639 29 15 ) ) PUNCT bracis-33639 29 16 are be AUX bracis-33639 29 17 combined combine VERB bracis-33639 29 18 to to PART bracis-33639 29 19 predict predict VERB bracis-33639 29 20 crime crime NOUN bracis-33639 29 21 hotspots hotspot NOUN bracis-33639 29 22 . . PUNCT bracis-33639 30 1 these these DET bracis-33639 30 2 studies study NOUN bracis-33639 30 3 demonstrate demonstrate VERB bracis-33639 30 4 the the DET bracis-33639 30 5 potential potential NOUN bracis-33639 30 6 of of ADP bracis-33639 30 7 gnns gnn NOUN bracis-33639 30 8 for for ADP bracis-33639 30 9 analyzing analyze VERB bracis-33639 30 10 and and CCONJ bracis-33639 30 11 forecasting forecast VERB bracis-33639 30 12 crime crime NOUN bracis-33639 30 13 events event NOUN bracis-33639 30 14 . . PUNCT bracis-33639 31 1 effective effective ADJ bracis-33639 31 2 data datum NOUN bracis-33639 31 3 modeling modeling NOUN bracis-33639 31 4 and and CCONJ bracis-33639 31 5 pre pre ADJ bracis-33639 31 6 - - ADJ bracis-33639 31 7 processing processing NOUN bracis-33639 31 8 are be AUX bracis-33639 31 9 crucial crucial ADJ bracis-33639 31 10 when when SCONJ bracis-33639 31 11 preparing prepare VERB bracis-33639 31 12 datasets dataset NOUN bracis-33639 31 13 for for ADP bracis-33639 31 14 graph graph NOUN bracis-33639 31 15 neural neural ADJ bracis-33639 31 16 networks network NOUN bracis-33639 31 17 , , PUNCT bracis-33639 31 18 particularly particularly ADV bracis-33639 31 19 for for ADP bracis-33639 31 20 crime crime NOUN bracis-33639 31 21 prediction prediction NOUN bracis-33639 31 22 tasks task NOUN bracis-33639 31 23 . . PUNCT bracis-33639 32 1 these these DET bracis-33639 32 2 processes process NOUN bracis-33639 32 3 dictate dictate VERB bracis-33639 32 4 how how SCONJ bracis-33639 32 5 raw raw ADJ bracis-33639 32 6 data datum NOUN bracis-33639 32 7 is be AUX bracis-33639 32 8 transformed transform VERB bracis-33639 32 9 and and CCONJ bracis-33639 32 10 integrated integrate VERB bracis-33639 32 11 into into ADP bracis-33639 32 12 graph graph NOUN bracis-33639 32 13 structures structure NOUN bracis-33639 32 14 , , PUNCT bracis-33639 32 15 significantly significantly ADV bracis-33639 32 16 impacting impact VERB bracis-33639 32 17 the the DET bracis-33639 32 18 model model NOUN bracis-33639 32 19 ’s ’s PART bracis-33639 32 20 performance performance NOUN bracis-33639 32 21 and and CCONJ bracis-33639 32 22 its its PRON bracis-33639 32 23 ability ability NOUN bracis-33639 32 24 to to PART bracis-33639 32 25 learn learn VERB bracis-33639 32 26 and and CCONJ bracis-33639 32 27 identify identify VERB bracis-33639 32 28 meaningful meaningful ADJ bracis-33639 32 29 patterns pattern NOUN bracis-33639 32 30 . . PUNCT bracis-33639 33 1 for for ADP bracis-33639 33 2 instance instance NOUN bracis-33639 33 3 , , PUNCT bracis-33639 33 4 when when SCONJ bracis-33639 33 5 street street NOUN bracis-33639 33 6 maps map NOUN bracis-33639 33 7 are be AUX bracis-33639 33 8 used use VERB bracis-33639 33 9 to to PART bracis-33639 33 10 generate generate VERB bracis-33639 33 11 graph graph NOUN bracis-33639 33 12 - - PUNCT bracis-33639 33 13 based base VERB bracis-33639 33 14 spatial spatial ADJ bracis-33639 33 15 discretization discretization NOUN bracis-33639 33 16 , , PUNCT bracis-33639 33 17 the the DET bracis-33639 33 18 nodes node NOUN bracis-33639 33 19 of of ADP bracis-33639 33 20 the the DET bracis-33639 33 21 graph graph NOUN bracis-33639 33 22 can can AUX bracis-33639 33 23 correspond correspond VERB bracis-33639 33 24 to to ADP bracis-33639 33 25 street street NOUN bracis-33639 33 26 segments segment NOUN bracis-33639 33 27 , , PUNCT bracis-33639 33 28 with with ADP bracis-33639 33 29 edges edge NOUN bracis-33639 33 30 connecting connect VERB bracis-33639 33 31 nodes node NOUN bracis-33639 33 32 whose whose DET bracis-33639 33 33 street street NOUN bracis-33639 33 34 segments segment NOUN bracis-33639 33 35 intersect intersect VERB bracis-33639 33 36 . . PUNCT bracis-33639 34 1 additionally additionally ADV bracis-33639 34 2 , , PUNCT bracis-33639 34 3 attributes attribute NOUN bracis-33639 34 4 can can AUX bracis-33639 34 5 be be AUX bracis-33639 34 6 assigned assign VERB bracis-33639 34 7 to to ADP bracis-33639 34 8 nodes node NOUN bracis-33639 34 9 and and CCONJ bracis-33639 34 10 edges edge NOUN bracis-33639 34 11 of of ADP bracis-33639 34 12 the the DET bracis-33639 34 13 graph graph NOUN bracis-33639 34 14 . . PUNCT bracis-33639 35 1 the the DET bracis-33639 35 2 challenge challenge NOUN bracis-33639 35 3 to to PART bracis-33639 35 4 make make VERB bracis-33639 35 5 this this DET bracis-33639 35 6 assignment assignment NOUN bracis-33639 35 7 lies lie VERB bracis-33639 35 8 in in ADP bracis-33639 35 9 integrating integrate VERB bracis-33639 35 10 data datum NOUN bracis-33639 35 11 from from ADP bracis-33639 35 12 different different ADJ bracis-33639 35 13 modalities modality NOUN bracis-33639 35 14 , , PUNCT bracis-33639 35 15 discretizations discretization NOUN bracis-33639 35 16 , , PUNCT bracis-33639 35 17 and and CCONJ bracis-33639 35 18 resolutions resolution NOUN bracis-33639 35 19 . . PUNCT bracis-33639 36 1 for for ADP bracis-33639 36 2 instance instance NOUN bracis-33639 36 3 , , PUNCT bracis-33639 36 4 socioeconomic socioeconomic ADJ bracis-33639 36 5 indicators indicator NOUN bracis-33639 36 6 , , PUNCT bracis-33639 36 7 population population NOUN bracis-33639 36 8 education education NOUN bracis-33639 36 9 levels level NOUN bracis-33639 36 10 , , PUNCT bracis-33639 36 11 and and CCONJ bracis-33639 36 12 the the DET bracis-33639 36 13 number number NOUN bracis-33639 36 14 of of ADP bracis-33639 36 15 points point NOUN bracis-33639 36 16 of of ADP bracis-33639 36 17 interest interest NOUN bracis-33639 36 18 ( ( PUNCT bracis-33639 36 19 such such ADJ bracis-33639 36 20 as as ADP bracis-33639 36 21 bars bar NOUN bracis-33639 36 22 , , PUNCT bracis-33639 36 23 bus bus NOUN bracis-33639 36 24 stops stop NOUN bracis-33639 36 25 , , PUNCT bracis-33639 36 26 and and CCONJ bracis-33639 36 27 schools school NOUN bracis-33639 36 28 ) ) PUNCT bracis-33639 36 29 near near ADP bracis-33639 36 30 each each DET bracis-33639 36 31 graph graph NOUN bracis-33639 36 32 node node NOUN bracis-33639 36 33 or or CCONJ bracis-33639 36 34 edge edge NOUN bracis-33639 36 35 are be AUX bracis-33639 36 36 typically typically ADV bracis-33639 36 37 discretized discretize VERB bracis-33639 36 38 into into ADP bracis-33639 36 39 census census NOUN bracis-33639 36 40 tracts tract NOUN bracis-33639 36 41 and and CCONJ bracis-33639 36 42 considered consider VERB bracis-33639 36 43 static static ADJ bracis-33639 36 44 information information NOUN bracis-33639 36 45 . . PUNCT bracis-33639 37 1 climate climate NOUN bracis-33639 37 2 data datum NOUN bracis-33639 37 3 is be AUX bracis-33639 37 4 provided provide VERB bracis-33639 37 5 at at ADP bracis-33639 37 6 a a DET bracis-33639 37 7 few few ADJ bracis-33639 37 8 specific specific ADJ bracis-33639 37 9 locations location NOUN bracis-33639 37 10 and and CCONJ bracis-33639 37 11 varies vary VERB bracis-33639 37 12 over over ADP bracis-33639 37 13 time time NOUN bracis-33639 37 14 , , PUNCT bracis-33639 37 15 while while SCONJ bracis-33639 37 16 crime crime NOUN bracis-33639 37 17 events event NOUN bracis-33639 37 18 are be AUX bracis-33639 37 19 geolocated geolocate VERB bracis-33639 37 20 and and CCONJ bracis-33639 37 21 typically typically ADV bracis-33639 37 22 dispersed disperse VERB bracis-33639 37 23 throughout throughout ADP bracis-33639 37 24 the the DET bracis-33639 37 25 spatial spatial ADJ bracis-33639 37 26 urban urban ADJ bracis-33639 37 27 domain domain NOUN bracis-33639 37 28 . . PUNCT bracis-33639 38 1 properly properly ADV bracis-33639 38 2 integrating integrate VERB bracis-33639 38 3 all all DET bracis-33639 38 4 these these DET bracis-33639 38 5 data datum NOUN bracis-33639 38 6 sources source NOUN bracis-33639 38 7 into into ADP bracis-33639 38 8 the the DET bracis-33639 38 9 nodes node NOUN bracis-33639 38 10 and and CCONJ bracis-33639 38 11 edges edge NOUN bracis-33639 38 12 of of ADP bracis-33639 38 13 the the DET bracis-33639 38 14 graph graph NOUN bracis-33639 38 15 is be AUX bracis-33639 38 16 a a DET bracis-33639 38 17 problem problem NOUN bracis-33639 38 18 for for ADP bracis-33639 38 19 which which PRON bracis-33639 38 20 no no DET bracis-33639 38 21 consolidated consolidated ADJ bracis-33639 38 22 methodology methodology NOUN bracis-33639 38 23 currently currently ADV bracis-33639 38 24 exists exist VERB bracis-33639 38 25 . . PUNCT bracis-33639 39 1 this this DET bracis-33639 39 2 work work NOUN bracis-33639 39 3 presents present VERB bracis-33639 39 4 a a DET bracis-33639 39 5 data data NOUN bracis-33639 39 6 modeling model VERB bracis-33639 39 7 methodology methodology NOUN bracis-33639 39 8 capable capable ADJ bracis-33639 39 9 of of ADP bracis-33639 39 10 integrating integrate VERB bracis-33639 39 11 data datum NOUN bracis-33639 39 12 from from ADP bracis-33639 39 13 different different ADJ bracis-33639 39 14 modalities modality NOUN bracis-33639 39 15 and and CCONJ bracis-33639 39 16 distinct distinct ADJ bracis-33639 39 17 discretization discretization NOUN bracis-33639 39 18 domains domain NOUN bracis-33639 39 19 into into ADP bracis-33639 39 20 a a DET bracis-33639 39 21 graph graph NOUN bracis-33639 39 22 structure structure NOUN bracis-33639 39 23 derived derive VERB bracis-33639 39 24 from from ADP bracis-33639 39 25 street street NOUN bracis-33639 39 26 maps map NOUN bracis-33639 39 27 . . PUNCT bracis-33639 40 1 the the DET bracis-33639 40 2 proposed propose VERB bracis-33639 40 3 data datum NOUN bracis-33639 40 4 modeling modeling NOUN bracis-33639 40 5 method method NOUN bracis-33639 40 6 has have AUX bracis-33639 40 7 been be AUX bracis-33639 40 8 employed employ VERB bracis-33639 40 9 to to ADP bracis-33639 40 10 structure structure NOUN bracis-33639 40 11 data datum NOUN bracis-33639 40 12 used use VERB bracis-33639 40 13 as as ADP bracis-33639 40 14 input input NOUN bracis-33639 40 15 for for ADP bracis-33639 40 16 two two NUM bracis-33639 40 17 distinct distinct ADJ bracis-33639 40 18 spatiotemporal spatiotemporal ADJ bracis-33639 40 19 gnn gnn NOUN bracis-33639 40 20 models model NOUN bracis-33639 40 21 . . PUNCT bracis-33639 41 1 specifically specifically ADV bracis-33639 41 2 , , PUNCT bracis-33639 41 3 we we PRON bracis-33639 41 4 evaluate evaluate VERB bracis-33639 41 5 and and CCONJ bracis-33639 41 6 compare compare VERB bracis-33639 41 7 the the DET bracis-33639 41 8 dynamic dynamic ADJ bracis-33639 41 9 self self NOUN bracis-33639 41 10 - - PUNCT bracis-33639 41 11 attention attention NOUN bracis-33639 41 12 network network NOUN bracis-33639 41 13 ( ( PUNCT bracis-33639 41 14 dysat dysat NOUN bracis-33639 41 15 ) ) PUNCT bracis-33639 42 1 [ [ X bracis-33639 42 2 26 26 NUM bracis-33639 42 3 ] ] PUNCT bracis-33639 42 4 and and CCONJ bracis-33639 42 5 the the DET bracis-33639 42 6 evolving evolve VERB bracis-33639 42 7 graph graph NOUN bracis-33639 42 8 convolutional convolutional ADJ bracis-33639 42 9 network network NOUN bracis-33639 42 10 ( ( PUNCT bracis-33639 42 11 evolvegcn evolvegcn PROPN bracis-33639 42 12 ) ) PUNCT bracis-33639 43 1 [ [ X bracis-33639 43 2 22 22 NUM bracis-33639 43 3 ] ] PUNCT bracis-33639 43 4 both both PRON bracis-33639 43 5 adapted adapt VERB bracis-33639 43 6 to to PART bracis-33639 43 7 operate operate VERB bracis-33639 43 8 in in ADP bracis-33639 43 9 crime crime NOUN bracis-33639 43 10 data datum NOUN bracis-33639 43 11 . . PUNCT bracis-33639 44 1 dysat dysat PROPN bracis-33639 44 2 employs employ VERB bracis-33639 44 3 a a DET bracis-33639 44 4 dynamic dynamic ADJ bracis-33639 44 5 self self NOUN bracis-33639 44 6 - - PUNCT bracis-33639 44 7 attention attention NOUN bracis-33639 44 8 mechanism mechanism NOUN bracis-33639 44 9 to to PART bracis-33639 44 10 model model VERB bracis-33639 44 11 the the DET bracis-33639 44 12 temporal temporal ADJ bracis-33639 44 13 evolution evolution NOUN bracis-33639 44 14 of of ADP bracis-33639 44 15 node node ADJ bracis-33639 44 16 representations representation NOUN bracis-33639 44 17 , , PUNCT bracis-33639 44 18 effectively effectively ADV bracis-33639 44 19 capturing capture VERB bracis-33639 44 20 long long ADJ bracis-33639 44 21 - - PUNCT bracis-33639 44 22 range range NOUN bracis-33639 44 23 dependencies dependency NOUN bracis-33639 44 24 in in ADP bracis-33639 44 25 both both DET bracis-33639 44 26 space space NOUN bracis-33639 44 27 and and CCONJ bracis-33639 44 28 time time NOUN bracis-33639 44 29 . . PUNCT bracis-33639 45 1 evolvegcn evolvegcn PROPN bracis-33639 45 2 , , PUNCT bracis-33639 45 3 on on ADP bracis-33639 45 4 the the DET bracis-33639 45 5 other other ADJ bracis-33639 45 6 hand hand NOUN bracis-33639 45 7 , , PUNCT bracis-33639 45 8 uses use VERB bracis-33639 45 9 recurrent recurrent ADJ bracis-33639 45 10 neural neural ADJ bracis-33639 45 11 network network NOUN bracis-33639 45 12 ( ( PUNCT bracis-33639 45 13 rnn rnn PROPN bracis-33639 45 14 ) ) PUNCT bracis-33639 45 15 architectures architecture NOUN bracis-33639 45 16 to to PART bracis-33639 45 17 dynamically dynamically ADV bracis-33639 45 18 update update VERB bracis-33639 45 19 the the DET bracis-33639 45 20 parameters parameter NOUN bracis-33639 45 21 of of ADP bracis-33639 45 22 the the DET bracis-33639 45 23 gcn gcn ADJ bracis-33639 45 24 layers layer NOUN bracis-33639 45 25 , , PUNCT bracis-33639 45 26 thus thus ADV bracis-33639 45 27 adapting adapt VERB bracis-33639 45 28 to to ADP bracis-33639 45 29 changes change NOUN bracis-33639 45 30 in in ADP bracis-33639 45 31 the the DET bracis-33639 45 32 graph graph NOUN bracis-33639 45 33 structure structure NOUN bracis-33639 45 34 over over ADP bracis-33639 45 35 time time NOUN bracis-33639 45 36 . . PUNCT bracis-33639 46 1 dysat dysat VERB bracis-33639 46 2 and and CCONJ bracis-33639 46 3 evolvegcn evolvegcn NOUN bracis-33639 46 4 performance performance NOUN bracis-33639 46 5 are be AUX bracis-33639 46 6 compared compare VERB bracis-33639 46 7 against against ADP bracis-33639 46 8 a a DET bracis-33639 46 9 baseline baseline NOUN bracis-33639 46 10 to to PART bracis-33639 46 11 better well ADV bracis-33639 46 12 understand understand VERB bracis-33639 46 13     SPACE bracis-33639 46 14 the the DET bracis-33639 46 15 advantages advantage NOUN bracis-33639 46 16 and and CCONJ bracis-33639 46 17 limitations limitation NOUN bracis-33639 46 18 of of ADP bracis-33639 46 19 the the DET bracis-33639 46 20 gnns gnns NOUN bracis-33639 46 21 in in ADP bracis-33639 46 22 the the DET bracis-33639 46 23 context context NOUN bracis-33639 46 24 of of ADP bracis-33639 46 25 crime crime NOUN bracis-33639 46 26 prediction prediction NOUN bracis-33639 46 27 . . PUNCT bracis-33639 47 1 our our PRON bracis-33639 47 2 contributions contribution NOUN bracis-33639 47 3 . . PUNCT bracis-33639 48 1 in in ADP bracis-33639 48 2 summary summary NOUN bracis-33639 48 3 , , PUNCT bracis-33639 48 4 the the DET bracis-33639 48 5 main main ADJ bracis-33639 48 6 contributions contribution NOUN bracis-33639 48 7 of of ADP bracis-33639 48 8 the the DET bracis-33639 48 9 present present ADJ bracis-33639 48 10 work work NOUN bracis-33639 48 11 are be AUX bracis-33639 48 12 : : PUNCT bracis-33639 48 13 1 1 X bracis-33639 48 14 . . X bracis-33639 48 15 data datum NOUN bracis-33639 48 16 modeling model VERB bracis-33639 48 17 pipeline pipeline NOUN bracis-33639 48 18 : : PUNCT bracis-33639 48 19 we we PRON bracis-33639 48 20 develop develop VERB bracis-33639 48 21 a a DET bracis-33639 48 22 robust robust ADJ bracis-33639 48 23 data datum NOUN bracis-33639 48 24 processing processing NOUN bracis-33639 48 25 pipeline pipeline NOUN bracis-33639 48 26 to to PART bracis-33639 48 27 integrate integrate VERB bracis-33639 48 28 crime crime NOUN bracis-33639 48 29 ( ( PUNCT bracis-33639 48 30 we we PRON bracis-33639 48 31 used use VERB bracis-33639 48 32 data datum NOUN bracis-33639 48 33 from from ADP bracis-33639 48 34 são são PROPN bracis-33639 48 35 paulo paulo PROPN bracis-33639 48 36 ’s ’s PART bracis-33639 48 37 department department PROPN bracis-33639 48 38 of of ADP bracis-33639 48 39 public public ADJ bracis-33639 48 40 safety safety NOUN bracis-33639 48 41 [ [ X bracis-33639 48 42 27 27 NUM bracis-33639 48 43 ] ] SYM bracis-33639 48 44 ) ) PUNCT bracis-33639 48 45 and and CCONJ bracis-33639 48 46 urban urban ADJ bracis-33639 48 47 infrastructure infrastructure NOUN bracis-33639 48 48 into into ADP bracis-33639 48 49 a a DET bracis-33639 48 50 graph graph NOUN bracis-33639 48 51 derived derive VERB bracis-33639 48 52 from from ADP bracis-33639 48 53 a a DET bracis-33639 48 54 street street NOUN bracis-33639 48 55 map map NOUN bracis-33639 48 56 . . PUNCT bracis-33639 49 1 this this DET bracis-33639 49 2 integration integration NOUN bracis-33639 49 3 allows allow VERB bracis-33639 49 4 for for ADP bracis-33639 49 5 a a DET bracis-33639 49 6 detailed detailed ADJ bracis-33639 49 7 and and CCONJ bracis-33639 49 8 dynamic dynamic ADJ bracis-33639 49 9 representation representation NOUN bracis-33639 49 10 of of ADP bracis-33639 49 11 the the DET bracis-33639 49 12 crime crime NOUN bracis-33639 49 13 landscape landscape NOUN bracis-33639 49 14 in in ADP bracis-33639 49 15 são são PROPN bracis-33639 49 16 paulo paulo PROPN bracis-33639 49 17 . . PUNCT bracis-33639 50 1 the the DET bracis-33639 50 2 proposed propose VERB bracis-33639 50 3 pipeline pipeline NOUN bracis-33639 50 4 is be AUX bracis-33639 50 5 versatile versatile ADV bracis-33639 50 6 enough enough ADV bracis-33639 50 7 to to PART bracis-33639 50 8 be be AUX bracis-33639 50 9 applied apply VERB bracis-33639 50 10 to to ADP bracis-33639 50 11 various various ADJ bracis-33639 50 12 domains domain NOUN bracis-33639 50 13 and and CCONJ bracis-33639 50 14 geographical geographical ADJ bracis-33639 50 15 areas area NOUN bracis-33639 50 16 , , PUNCT bracis-33639 50 17 requiring require VERB bracis-33639 50 18 only only ADV bracis-33639 50 19 the the DET bracis-33639 50 20 data datum NOUN bracis-33639 50 21 is be AUX bracis-33639 50 22 geolocated geolocate VERB bracis-33639 50 23 . . PUNCT bracis-33639 51 1 2 2 X bracis-33639 51 2 . . X bracis-33639 51 3 gnns gnns PROPN bracis-33639 51 4 for for ADP bracis-33639 51 5 crime crime NOUN bracis-33639 51 6 data datum NOUN bracis-33639 51 7 : : PUNCT bracis-33639 51 8 we we PRON bracis-33639 51 9 adapt adapt VERB bracis-33639 51 10 two two NUM bracis-33639 51 11 distinct distinct ADJ bracis-33639 51 12 spatiotemporal spatiotemporal ADJ bracis-33639 51 13 gnn gnn PROPN bracis-33639 51 14 architectures architecture NOUN bracis-33639 51 15 , , PUNCT bracis-33639 51 16 dysat dysat ADJ bracis-33639 51 17 and and CCONJ bracis-33639 51 18 evolvegcn evolvegcn NOUN bracis-33639 51 19 , , PUNCT bracis-33639 51 20 for for ADP bracis-33639 51 21 crime crime NOUN bracis-33639 51 22 prediction prediction NOUN bracis-33639 51 23 . . PUNCT bracis-33639 52 1 to to ADP bracis-33639 52 2 our our PRON bracis-33639 52 3 knowledge knowledge NOUN bracis-33639 52 4 , , PUNCT bracis-33639 52 5 these these DET bracis-33639 52 6 models model NOUN bracis-33639 52 7 have have AUX bracis-33639 52 8 not not PART bracis-33639 52 9 been be AUX bracis-33639 52 10 previously previously ADV bracis-33639 52 11 applied apply VERB bracis-33639 52 12 to to ADP bracis-33639 52 13 the the DET bracis-33639 52 14 context context NOUN bracis-33639 52 15 of of ADP bracis-33639 52 16 crime crime NOUN bracis-33639 52 17 prediction prediction NOUN bracis-33639 52 18 . . PUNCT bracis-33639 53 1 3 3 X bracis-33639 53 2 . . NOUN bracis-33639 53 3 models model NOUN bracis-33639 53 4 evaluation evaluation NOUN bracis-33639 53 5 and and CCONJ bracis-33639 53 6 comparison comparison NOUN bracis-33639 53 7 : : PUNCT bracis-33639 53 8 we we PRON bracis-33639 53 9 thoroughly thoroughly ADV bracis-33639 53 10 and and CCONJ bracis-33639 53 11 comprehensively comprehensively ADV bracis-33639 53 12 evaluate evaluate VERB bracis-33639 53 13 and and CCONJ bracis-33639 53 14 compare compare VERB bracis-33639 53 15 dysat dysat NOUN bracis-33639 53 16 and and CCONJ bracis-33639 53 17 evolvegcn evolvegcn NOUN bracis-33639 53 18 using use VERB bracis-33639 53 19 the the DET bracis-33639 53 20 processed process VERB bracis-33639 53 21 são são PROPN bracis-33639 53 22 paulo paulo PROPN bracis-33639 53 23 crime crime PROPN bracis-33639 53 24 data data PROPN bracis-33639 53 25 , , PUNCT bracis-33639 53 26 assessing assess VERB bracis-33639 53 27 their their PRON bracis-33639 53 28 effectiveness effectiveness NOUN bracis-33639 53 29 in in ADP bracis-33639 53 30 learning learn VERB bracis-33639 53 31 from from ADP bracis-33639 53 32 historical historical ADJ bracis-33639 53 33 data datum NOUN bracis-33639 53 34 and and CCONJ bracis-33639 53 35 predicting predict VERB bracis-33639 53 36 future future ADJ bracis-33639 53 37 crime crime NOUN bracis-33639 53 38 occurrences occurrence NOUN bracis-33639 53 39 . . PUNCT bracis-33639 54 1 this this DET bracis-33639 54 2 study study NOUN bracis-33639 54 3 explores explore VERB bracis-33639 54 4 the the DET bracis-33639 54 5 effectiveness effectiveness NOUN bracis-33639 54 6 of of ADP bracis-33639 54 7 gnn gnn PROPN bracis-33639 54 8 architectures architecture NOUN bracis-33639 54 9 , , PUNCT bracis-33639 54 10 specifically specifically ADV bracis-33639 54 11 dysat dysat VERB bracis-33639 54 12 and and CCONJ bracis-33639 54 13 evolvegcn evolvegcn ADJ bracis-33639 54 14 , , PUNCT bracis-33639 54 15 in in ADP bracis-33639 54 16 the the DET bracis-33639 54 17 crime crime NOUN bracis-33639 54 18 domain domain NOUN bracis-33639 54 19 by by ADP bracis-33639 54 20 applying apply VERB bracis-33639 54 21 , , PUNCT bracis-33639 54 22 evaluating evaluating NOUN bracis-33639 54 23 , , PUNCT bracis-33639 54 24 and and CCONJ bracis-33639 54 25 comparing compare VERB bracis-33639 54 26 these these DET bracis-33639 54 27 models model NOUN bracis-33639 54 28 on on ADP bracis-33639 54 29 crime crime NOUN bracis-33639 54 30 data datum NOUN bracis-33639 54 31 using use VERB bracis-33639 54 32 our our PRON bracis-33639 54 33 developed develop VERB bracis-33639 54 34 data data NOUN bracis-33639 54 35 pipeline pipeline NOUN bracis-33639 54 36 . . PUNCT bracis-33639 55 1 integrating integrate VERB bracis-33639 55 2 street street NOUN bracis-33639 55 3 maps map NOUN bracis-33639 55 4 and and CCONJ bracis-33639 55 5 crime crime NOUN bracis-33639 55 6 data datum NOUN bracis-33639 55 7 , , PUNCT bracis-33639 55 8 combined combine VERB bracis-33639 55 9 with with ADP bracis-33639 55 10 advanced advanced ADJ bracis-33639 55 11 spatiotemporal spatiotemporal ADJ bracis-33639 55 12 modeling modeling NOUN bracis-33639 55 13 techniques technique NOUN bracis-33639 55 14 , , PUNCT bracis-33639 55 15 provides provide VERB bracis-33639 55 16 a a DET bracis-33639 55 17 novel novel ADJ bracis-33639 55 18 approach approach NOUN bracis-33639 55 19 to to ADP bracis-33639 55 20 understanding understanding NOUN bracis-33639 55 21 and and CCONJ bracis-33639 55 22 predicting predict VERB bracis-33639 55 23 crime crime NOUN bracis-33639 55 24 dynamics dynamic NOUN bracis-33639 55 25 in in ADP bracis-33639 55 26 urban urban ADJ bracis-33639 55 27 environments environment NOUN bracis-33639 55 28 . . PUNCT bracis-33639 56 1 our our PRON bracis-33639 56 2 findings finding NOUN bracis-33639 56 3 offer offer VERB bracis-33639 56 4 valuable valuable ADJ bracis-33639 56 5 insights insight NOUN bracis-33639 56 6 for for ADP bracis-33639 56 7 researchers researcher NOUN bracis-33639 56 8 and and CCONJ bracis-33639 56 9 practitioners practitioner NOUN bracis-33639 56 10 working work VERB bracis-33639 56 11 on on ADP bracis-33639 56 12 crime crime NOUN bracis-33639 56 13 prevention prevention NOUN bracis-33639 56 14 and and CCONJ bracis-33639 56 15 public public ADJ bracis-33639 56 16 safety safety NOUN bracis-33639 56 17 enhancement enhancement NOUN bracis-33639 56 18 . . PUNCT bracis-33639 57 1 2 2 NUM bracis-33639 57 2 related relate VERB bracis-33639 57 3 work work NOUN bracis-33639 57 4 crime crime NOUN bracis-33639 57 5 analytics analytic NOUN bracis-33639 57 6 has have AUX bracis-33639 57 7 been be AUX bracis-33639 57 8 an an DET bracis-33639 57 9 essential essential ADJ bracis-33639 57 10 area area NOUN bracis-33639 57 11 of of ADP bracis-33639 57 12 urban urban ADJ bracis-33639 57 13 safety safety NOUN bracis-33639 57 14 and and CCONJ bracis-33639 57 15 policy policy NOUN bracis-33639 57 16 - - PUNCT bracis-33639 57 17 making make VERB bracis-33639 57 18 study study NOUN bracis-33639 57 19 . . PUNCT bracis-33639 58 1 traditional traditional ADJ bracis-33639 58 2 studies study NOUN bracis-33639 58 3 primarily primarily ADV bracis-33639 58 4 focus focus VERB bracis-33639 58 5 on on ADP bracis-33639 58 6 statistical statistical ADJ bracis-33639 58 7 methods method NOUN bracis-33639 58 8 and and CCONJ bracis-33639 58 9 classical classical ADJ bracis-33639 58 10 machine machine NOUN bracis-33639 58 11 learning learning NOUN bracis-33639 58 12 models model NOUN bracis-33639 58 13 . . PUNCT bracis-33639 59 1 for for ADP bracis-33639 59 2 instance instance NOUN bracis-33639 59 3 , , PUNCT bracis-33639 59 4 gorr gorr NOUN bracis-33639 59 5 and and CCONJ bracis-33639 59 6 harries harrie NOUN bracis-33639 59 7     SPACE bracis-33639 59 8 [ [ X bracis-33639 59 9 9 9 NUM bracis-33639 59 10 ] ] PUNCT bracis-33639 59 11 employed employ VERB bracis-33639 59 12 time time NOUN bracis-33639 59 13 - - PUNCT bracis-33639 59 14 series series NOUN bracis-33639 59 15 regression regression NOUN bracis-33639 59 16 models model NOUN bracis-33639 59 17 to to PART bracis-33639 59 18 forecast forecast VERB bracis-33639 59 19 crime crime NOUN bracis-33639 59 20 rates rate NOUN bracis-33639 59 21 , , PUNCT bracis-33639 59 22 concentrating concentrate VERB bracis-33639 59 23 primarily primarily ADV bracis-33639 59 24 on on ADP bracis-33639 59 25 the the DET bracis-33639 59 26 temporal temporal ADJ bracis-33639 59 27 elements element NOUN bracis-33639 59 28 of of ADP bracis-33639 59 29 crime crime NOUN bracis-33639 59 30 data datum NOUN bracis-33639 59 31 . . PUNCT bracis-33639 60 1 these these DET bracis-33639 60 2 models model NOUN bracis-33639 60 3 , , PUNCT bracis-33639 60 4 while while SCONJ bracis-33639 60 5 valuable valuable ADJ bracis-33639 60 6 , , PUNCT bracis-33639 60 7 often often ADV bracis-33639 60 8 needed need VERB bracis-33639 60 9 to to PART bracis-33639 60 10 account account VERB bracis-33639 60 11 for for ADP bracis-33639 60 12 the the DET bracis-33639 60 13 spatial spatial ADJ bracis-33639 60 14 dependencies dependency NOUN bracis-33639 60 15 that that PRON bracis-33639 60 16 inherently inherently ADV bracis-33639 60 17 exist exist VERB bracis-33639 60 18 in in ADP bracis-33639 60 19 crime crime NOUN bracis-33639 60 20 patterns pattern NOUN bracis-33639 60 21 . . PUNCT bracis-33639 61 1 spatial spatial ADJ bracis-33639 61 2 analysis analysis NOUN bracis-33639 61 3 techniques technique NOUN bracis-33639 61 4 , , PUNCT bracis-33639 61 5 such such ADJ bracis-33639 61 6 as as ADP bracis-33639 61 7 hotspot hotspot ADJ bracis-33639 61 8 analysis analysis NOUN bracis-33639 61 9 and and CCONJ bracis-33639 61 10 spatial spatial ADJ bracis-33639 61 11 clustering clustering NOUN bracis-33639 61 12 [ [ X bracis-33639 61 13 6 6 NUM bracis-33639 61 14 ] ] PUNCT bracis-33639 61 15 , , PUNCT bracis-33639 61 16 have have AUX bracis-33639 61 17 been be AUX bracis-33639 61 18 used use VERB bracis-33639 61 19 to to PART bracis-33639 61 20 identify identify VERB bracis-33639 61 21 areas area NOUN bracis-33639 61 22 with with ADP bracis-33639 61 23 high high ADJ bracis-33639 61 24 crime crime NOUN bracis-33639 61 25 rates rate NOUN bracis-33639 61 26 . . PUNCT bracis-33639 62 1 these these DET bracis-33639 62 2 methods method NOUN bracis-33639 62 3 highlight highlight VERB bracis-33639 62 4 the the DET bracis-33639 62 5 geographic geographic ADJ bracis-33639 62 6 concentration concentration NOUN bracis-33639 62 7 of of ADP bracis-33639 62 8 crimes crime NOUN bracis-33639 62 9 but but CCONJ bracis-33639 62 10 typically typically ADV bracis-33639 62 11 lack lack VERB bracis-33639 62 12 the the DET bracis-33639 62 13 temporal temporal ADJ bracis-33639 62 14 dimension dimension NOUN bracis-33639 62 15 essential essential ADJ bracis-33639 62 16 for for ADP bracis-33639 62 17 dynamic dynamic ADJ bracis-33639 62 18 crime crime NOUN bracis-33639 62 19 prediction prediction NOUN bracis-33639 62 20 . . PUNCT bracis-33639 63 1 recent recent ADJ bracis-33639 63 2 advancements advancement NOUN bracis-33639 63 3 in in ADP bracis-33639 63 4 spatio spatio PROPN bracis-33639 63 5 - - PUNCT bracis-33639 63 6 temporal temporal ADJ bracis-33639 63 7 data datum NOUN bracis-33639 63 8 modeling modeling NOUN bracis-33639 63 9 have have AUX bracis-33639 63 10 enabled enable VERB bracis-33639 63 11 the the DET bracis-33639 63 12 development development NOUN bracis-33639 63 13 of of ADP bracis-33639 63 14 sophisticated sophisticated ADJ bracis-33639 63 15 and and CCONJ bracis-33639 63 16 complex complex ADJ bracis-33639 63 17 analytical analytical ADJ bracis-33639 63 18 tools tool NOUN bracis-33639 63 19 for for ADP bracis-33639 63 20 analyzing analyze VERB bracis-33639 63 21 complex complex ADJ bracis-33639 63 22 phenomena phenomenon NOUN bracis-33639 63 23 [ [ X bracis-33639 63 24 8 8 NUM bracis-33639 63 25 ] ] PUNCT bracis-33639 63 26 . . PUNCT bracis-33639 64 1 for for ADP bracis-33639 64 2 instance instance NOUN bracis-33639 64 3 , , PUNCT bracis-33639 64 4 salinas salinas PROPN bracis-33639 64 5 et et PROPN bracis-33639 64 6 al al PROPN bracis-33639 64 7 . . PUNCT bracis-33639 64 8     SPACE bracis-33639 65 1 [ [ X bracis-33639 65 2 25 25 NUM bracis-33639 65 3 ] ] PUNCT bracis-33639 65 4 proposed propose VERB bracis-33639 65 5 a a DET bracis-33639 65 6 city city NOUN bracis-33639 65 7 hub hub NOUN bracis-33639 65 8 framework framework NOUN bracis-33639 65 9 that that PRON bracis-33639 65 10 integrates integrate VERB bracis-33639 65 11 data datum NOUN bracis-33639 65 12 from from ADP bracis-33639 65 13 different different ADJ bracis-33639 65 14 sources source NOUN bracis-33639 65 15 on on ADP bracis-33639 65 16 the the DET bracis-33639 65 17 nodes node NOUN bracis-33639 65 18 of of ADP bracis-33639 65 19 a a DET bracis-33639 65 20 street street NOUN bracis-33639 65 21 map map NOUN bracis-33639 65 22 graph graph NOUN bracis-33639 65 23 , , PUNCT bracis-33639 65 24 aiming aim VERB bracis-33639 65 25 mainly mainly ADV bracis-33639 65 26 to to PART bracis-33639 65 27 investigate investigate VERB bracis-33639 65 28 crime crime NOUN bracis-33639 65 29 - - PUNCT bracis-33639 65 30 related relate VERB bracis-33639 65 31 phenomena phenomenon NOUN bracis-33639 65 32 . . PUNCT bracis-33639 66 1 although although SCONJ bracis-33639 66 2 their their PRON bracis-33639 66 3 goal goal NOUN bracis-33639 66 4 was be AUX bracis-33639 66 5 more more ADV bracis-33639 66 6 on on ADP bracis-33639 66 7 visualizing visualize VERB bracis-33639 66 8 and and CCONJ bracis-33639 66 9 analyzing analyze VERB bracis-33639 66 10 crime crime NOUN bracis-33639 66 11 events event NOUN bracis-33639 66 12 , , PUNCT bracis-33639 66 13 they they PRON bracis-33639 66 14 leveraged leveraged ADJ bracis-33639 66 15 crime crime NOUN bracis-33639 66 16 prediction prediction NOUN bracis-33639 66 17 models model NOUN bracis-33639 66 18 to to PART bracis-33639 66 19 support support VERB bracis-33639 66 20 the the DET bracis-33639 66 21 analysis analysis NOUN bracis-33639 66 22 . . PUNCT bracis-33639 67 1 to to PART bracis-33639 67 2 analyze analyze VERB bracis-33639 67 3 human human ADJ bracis-33639 67 4 and and CCONJ bracis-33639 67 5 drug drug NOUN bracis-33639 67 6 trafficking trafficking NOUN bracis-33639 67 7 crimes crime NOUN bracis-33639 67 8 ahmed ahme VERB bracis-33639 67 9 et et PROPN bracis-33639 67 10 al al PROPN bracis-33639 67 11 . . PUNCT bracis-33639 67 12     SPACE bracis-33639 68 1 [ [ X bracis-33639 68 2 2 2 X bracis-33639 68 3 ] ] PUNCT bracis-33639 68 4 used use VERB bracis-33639 68 5 entity entity NOUN bracis-33639 68 6 resolution resolution NOUN bracis-33639 68 7 techniques technique NOUN bracis-33639 68 8 to to PART bracis-33639 68 9 merge merge VERB bracis-33639 68 10 multiple multiple ADJ bracis-33639 68 11 state state NOUN bracis-33639 68 12 - - PUNCT bracis-33639 68 13 wide wide ADJ bracis-33639 68 14 and and CCONJ bracis-33639 68 15 county county NOUN bracis-33639 68 16 - - PUNCT bracis-33639 68 17 wide wide ADJ bracis-33639 68 18 crime crime NOUN bracis-33639 68 19 datasets dataset NOUN bracis-33639 68 20 into into ADP bracis-33639 68 21 a a DET bracis-33639 68 22 geographic geographic ADJ bracis-33639 68 23 graph graph NOUN bracis-33639 68 24 , , PUNCT bracis-33639 68 25 combining combine VERB bracis-33639 68 26 incident incident NOUN bracis-33639 68 27 reports report NOUN bracis-33639 68 28 , , PUNCT bracis-33639 68 29 crime crime NOUN bracis-33639 68 30 reports report NOUN bracis-33639 68 31 , , PUNCT bracis-33639 68 32 and and CCONJ bracis-33639 68 33 court court NOUN bracis-33639 68 34 records record NOUN bracis-33639 68 35 , , PUNCT bracis-33639 68 36 in in ADP bracis-33639 68 37 a a DET bracis-33639 68 38 single single ADJ bracis-33639 68 39 dataset dataset NOUN bracis-33639 68 40 . . PUNCT bracis-33639 69 1 however however ADV bracis-33639 69 2 , , PUNCT bracis-33639 69 3 their their PRON bracis-33639 69 4 data datum NOUN bracis-33639 69 5 do do AUX bracis-33639 69 6 not not PART bracis-33639 69 7 account account VERB bracis-33639 69 8 for for ADP bracis-33639 69 9 urban urban ADJ bracis-33639 69 10 - - PUNCT bracis-33639 69 11 related relate VERB bracis-33639 69 12 variables variable NOUN bracis-33639 69 13 , , PUNCT bracis-33639 69 14 which which PRON bracis-33639 69 15 are be AUX bracis-33639 69 16 fundamental fundamental ADJ bracis-33639 69 17 for for ADP bracis-33639 69 18 crime crime NOUN bracis-33639 69 19 analysis analysis NOUN bracis-33639 69 20 . . PUNCT bracis-33639 70 1 the the DET bracis-33639 70 2 work work NOUN bracis-33639 70 3 by by ADP bracis-33639 70 4 salinas salinas PROPN bracis-33639 70 5 et et PROPN bracis-33639 70 6 al al PROPN bracis-33639 70 7 . . PUNCT bracis-33639 70 8     SPACE bracis-33639 71 1 [ [ X bracis-33639 71 2 25 25 NUM bracis-33639 71 3 ] ] PUNCT bracis-33639 71 4 and and CCONJ bracis-33639 71 5 ahmed ahmed PROPN bracis-33639 71 6 et et PROPN bracis-33639 71 7 al al PROPN bracis-33639 71 8 . . PUNCT bracis-33639 71 9     SPACE bracis-33639 72 1 [ [ X bracis-33639 72 2 2 2 NUM bracis-33639 72 3 ] ] PUNCT bracis-33639 72 4 are be AUX bracis-33639 72 5 examples example NOUN bracis-33639 72 6 of of ADP bracis-33639 72 7 methods method NOUN bracis-33639 72 8 designed design VERB bracis-33639 72 9 to to PART bracis-33639 72 10 integrate integrate VERB bracis-33639 72 11 data datum NOUN bracis-33639 72 12 into into ADP bracis-33639 72 13 a a DET bracis-33639 72 14 graph graph NOUN bracis-33639 72 15 structure structure NOUN bracis-33639 72 16 . . PUNCT bracis-33639 73 1 however however ADV bracis-33639 73 2 , , PUNCT bracis-33639 73 3 as as ADV bracis-33639 73 4 far far ADV bracis-33639 73 5 as as SCONJ bracis-33639 73 6 we we PRON bracis-33639 73 7 know know VERB bracis-33639 73 8 , , PUNCT bracis-33639 73 9 there there PRON bracis-33639 73 10 is be VERB bracis-33639 73 11 no no DET bracis-33639 73 12 consolidated consolidated ADJ bracis-33639 73 13 methodology methodology NOUN bracis-33639 73 14 to to PART bracis-33639 73 15 perform perform VERB bracis-33639 73 16 such such DET bracis-33639 73 17 an an DET bracis-33639 73 18 integration integration NOUN bracis-33639 73 19 . . PUNCT bracis-33639 74 1 deep deep ADJ bracis-33639 74 2 learning learning NOUN bracis-33639 74 3 models model NOUN bracis-33639 74 4 such such ADJ bracis-33639 74 5 as as ADP bracis-33639 74 6 convolutional convolutional ADJ bracis-33639 74 7 neural neural ADJ bracis-33639 74 8 networks network NOUN bracis-33639 74 9 [ [ X bracis-33639 74 10 21 21 NUM bracis-33639 74 11 ] ] PUNCT bracis-33639 74 12 and and CCONJ bracis-33639 74 13 recurrent recurrent ADJ bracis-33639 74 14 neural neural ADJ bracis-33639 74 15 networks network NOUN bracis-33639 74 16 [ [ X bracis-33639 74 17 19 19 NUM bracis-33639 74 18 ] ] PUNCT bracis-33639 74 19 have have AUX bracis-33639 74 20 been be AUX bracis-33639 74 21 adapted adapt VERB bracis-33639 74 22 for for ADP bracis-33639 74 23 handling handle VERB bracis-33639 74 24 spatio spatio PROPN bracis-33639 74 25 - - PUNCT bracis-33639 74 26 temporal temporal ADJ bracis-33639 74 27 data datum NOUN bracis-33639 74 28 , , PUNCT bracis-33639 74 29 with with ADP bracis-33639 74 30 noteworthy noteworthy ADJ bracis-33639 74 31 success success NOUN bracis-33639 74 32 . . PUNCT bracis-33639 75 1 shi shi PROPN bracis-33639 75 2 at at ADP bracis-33639 75 3 al al PROPN bracis-33639 75 4 . . PUNCT bracis-33639 75 5     SPACE bracis-33639 76 1 [ [ X bracis-33639 76 2 28 28 NUM bracis-33639 76 3 ] ] PUNCT bracis-33639 76 4 , , PUNCT bracis-33639 76 5 for for ADP bracis-33639 76 6 instance instance NOUN bracis-33639 76 7 , , PUNCT bracis-33639 76 8 demonstrated demonstrate VERB bracis-33639 76 9 the the DET bracis-33639 76 10 effectiveness effectiveness NOUN bracis-33639 76 11 of of ADP bracis-33639 76 12 cnns cnn NOUN bracis-33639 76 13 for for ADP bracis-33639 76 14 spatial spatial ADJ bracis-33639 76 15 feature feature NOUN bracis-33639 76 16 extraction extraction NOUN bracis-33639 76 17 and and CCONJ bracis-33639 76 18 rnns rnn NOUN bracis-33639 76 19 for for ADP bracis-33639 76 20 capturing capture VERB bracis-33639 76 21 temporal temporal ADJ bracis-33639 76 22 dependencies dependency NOUN bracis-33639 76 23 in in ADP bracis-33639 76 24 video video NOUN bracis-33639 76 25 data datum NOUN bracis-33639 76 26 . . PUNCT bracis-33639 77 1 despite despite SCONJ bracis-33639 77 2 these these DET bracis-33639 77 3 advancements advancement NOUN bracis-33639 77 4 , , PUNCT bracis-33639 77 5 these these DET bracis-33639 77 6 models model NOUN bracis-33639 77 7 frequently frequently ADV bracis-33639 77 8 struggle struggle VERB bracis-33639 77 9 with with ADP bracis-33639 77 10 irregular irregular ADJ bracis-33639 77 11 and and CCONJ bracis-33639 77 12 non non ADJ bracis-33639 77 13 - - ADJ bracis-33639 77 14 euclidean euclidean ADJ bracis-33639 77 15 data datum NOUN bracis-33639 77 16 structures structure NOUN bracis-33639 77 17 , , PUNCT bracis-33639 77 18 which which PRON bracis-33639 77 19 are be AUX bracis-33639 77 20 natural natural ADJ bracis-33639 77 21 real real ADJ bracis-33639 77 22 - - PUNCT bracis-33639 77 23 world world NOUN bracis-33639 77 24 representations representation NOUN bracis-33639 77 25 , , PUNCT bracis-33639 77 26 particularly particularly ADV bracis-33639 77 27 for for ADP bracis-33639 77 28 urban urban ADJ bracis-33639 77 29 analytics analytic NOUN bracis-33639 77 30     SPACE bracis-33639 78 1 [ [ X bracis-33639 78 2 4 4 NUM bracis-33639 78 3 ] ] PUNCT bracis-33639 78 4 . . PUNCT bracis-33639 79 1 graph graph NOUN bracis-33639 79 2 neural neural ADJ bracis-33639 79 3 networks network NOUN bracis-33639 79 4 ( ( PUNCT bracis-33639 79 5 gnns gnns NOUN bracis-33639 79 6 ) ) PUNCT bracis-33639 79 7 have have AUX bracis-33639 79 8 emerged emerge VERB bracis-33639 79 9 as as ADP bracis-33639 79 10 robust robust ADJ bracis-33639 79 11 tools tool NOUN bracis-33639 79 12 for for ADP bracis-33639 79 13 modeling model VERB bracis-33639 79 14 complex complex ADJ bracis-33639 79 15 graph graph NOUN bracis-33639 79 16 - - PUNCT bracis-33639 79 17 structured structure VERB bracis-33639 79 18 data datum NOUN bracis-33639 79 19 . . PUNCT bracis-33639 80 1 the the DET bracis-33639 80 2 work work NOUN bracis-33639 80 3 by by ADP bracis-33639 80 4 kipf kipf VERB bracis-33639 80 5 and and CCONJ bracis-33639 80 6 welling well VERB bracis-33639 80 7     SPACE bracis-33639 80 8 [ [ X bracis-33639 80 9 15 15 NUM bracis-33639 80 10 ] ] PUNCT bracis-33639 80 11 on on ADP bracis-33639 80 12 graph graph NOUN bracis-33639 80 13 convolutional convolutional ADJ bracis-33639 80 14 networks network NOUN bracis-33639 80 15 ( ( PUNCT bracis-33639 80 16 gcns gcns PROPN bracis-33639 80 17 ) ) PUNCT bracis-33639 80 18 indicated indicate VERB bracis-33639 80 19 that that SCONJ bracis-33639 80 20 gnns gnns NOUN bracis-33639 80 21 can can AUX bracis-33639 80 22 leverage leverage VERB bracis-33639 80 23 connectivity connectivity NOUN bracis-33639 80 24 information information NOUN bracis-33639 80 25 present present ADJ bracis-33639 80 26 in in ADP bracis-33639 80 27 graph graph NOUN bracis-33639 80 28 data datum NOUN bracis-33639 80 29 to to PART bracis-33639 80 30 improve improve VERB bracis-33639 80 31 predictive predictive ADJ bracis-33639 80 32 performance performance NOUN bracis-33639 80 33 . . PUNCT bracis-33639 81 1 however however ADV bracis-33639 81 2 , , PUNCT bracis-33639 81 3 static static ADJ bracis-33639 81 4 gnns gnn NOUN bracis-33639 81 5 are be AUX bracis-33639 81 6 limited limit VERB bracis-33639 81 7 in in ADP bracis-33639 81 8 their their PRON bracis-33639 81 9 ability ability NOUN bracis-33639 81 10 to to PART bracis-33639 81 11 model model VERB bracis-33639 81 12 temporal temporal ADJ bracis-33639 81 13 dynamics dynamic NOUN bracis-33639 81 14 , , PUNCT bracis-33639 81 15 which which PRON bracis-33639 81 16 is be AUX bracis-33639 81 17 critical critical ADJ bracis-33639 81 18 for for ADP bracis-33639 81 19 many many ADJ bracis-33639 81 20 real real ADJ bracis-33639 81 21 - - PUNCT bracis-33639 81 22 world world NOUN bracis-33639 81 23 applications application NOUN bracis-33639 81 24 , , PUNCT bracis-33639 81 25 including include VERB bracis-33639 81 26 crime crime NOUN bracis-33639 81 27 prediction prediction NOUN bracis-33639 81 28 . . PUNCT bracis-33639 82 1 dynamic dynamic ADJ bracis-33639 82 2 gnns gnn NOUN bracis-33639 82 3 extend extend VERB bracis-33639 82 4 static static ADJ bracis-33639 82 5 gnns gnn NOUN bracis-33639 82 6 by by ADP bracis-33639 82 7 integrating integrate VERB bracis-33639 82 8 temporal temporal ADJ bracis-33639 82 9 information information NOUN bracis-33639 82 10 , , PUNCT bracis-33639 82 11 making make VERB bracis-33639 82 12 them they PRON bracis-33639 82 13 suitable suitable ADJ bracis-33639 82 14 for for ADP bracis-33639 82 15 spatio spatio PROPN bracis-33639 82 16 - - PUNCT bracis-33639 82 17 temporal temporal ADJ bracis-33639 82 18 prediction prediction NOUN bracis-33639 82 19 tasks task NOUN bracis-33639 82 20 . . PUNCT bracis-33639 83 1 for for ADP bracis-33639 83 2 instance instance NOUN bracis-33639 83 3 , , PUNCT bracis-33639 83 4 the the DET bracis-33639 83 5 dynamic dynamic ADJ bracis-33639 83 6 self self NOUN bracis-33639 83 7 - - PUNCT bracis-33639 83 8 attention attention NOUN bracis-33639 83 9 network network NOUN bracis-33639 83 10 ( ( PUNCT bracis-33639 83 11 dysat dysat NOUN bracis-33639 83 12 ) ) PUNCT bracis-33639 84 1 [ [ X bracis-33639 84 2 26 26 NUM bracis-33639 84 3 ] ] PUNCT bracis-33639 84 4 employs employ VERB bracis-33639 84 5 a a DET bracis-33639 84 6 self self NOUN bracis-33639 84 7 - - PUNCT bracis-33639 84 8 attention attention NOUN bracis-33639 84 9 mechanism mechanism NOUN bracis-33639 84 10 to to PART bracis-33639 84 11 capture capture VERB bracis-33639 84 12 temporal temporal ADJ bracis-33639 84 13 dependencies dependency NOUN bracis-33639 84 14 within within ADP bracis-33639 84 15 dynamic dynamic ADJ bracis-33639 84 16 graphs graph NOUN bracis-33639 84 17 . . PUNCT bracis-33639 85 1 similarly similarly ADV bracis-33639 85 2 , , PUNCT bracis-33639 85 3 evolving evolve VERB bracis-33639 85 4 graph graph NOUN bracis-33639 85 5 convolutional convolutional ADJ bracis-33639 85 6 networks network NOUN bracis-33639 85 7 ( ( PUNCT bracis-33639 85 8 evolvegcn evolvegcn PROPN bracis-33639 85 9 ) ) PUNCT bracis-33639 86 1 [ [ X bracis-33639 86 2 22 22 NUM bracis-33639 86 3 ] ] PUNCT bracis-33639 86 4 use use VERB bracis-33639 86 5 recurrent recurrent ADJ bracis-33639 86 6 neural neural ADJ bracis-33639 86 7 networks network NOUN bracis-33639 86 8 to to PART bracis-33639 86 9 update update VERB bracis-33639 86 10 node node ADJ bracis-33639 86 11 embeddings embedding NOUN bracis-33639 86 12 over over ADP bracis-33639 86 13 time time NOUN bracis-33639 86 14 , , PUNCT bracis-33639 86 15 effectively effectively ADV bracis-33639 86 16 capturing capture VERB bracis-33639 86 17 the the DET bracis-33639 86 18 evolution evolution NOUN bracis-33639 86 19 of of ADP bracis-33639 86 20 graph graph NOUN bracis-33639 86 21 structures structure NOUN bracis-33639 86 22 . . PUNCT bracis-33639 87 1 these these DET bracis-33639 87 2 dynamic dynamic ADJ bracis-33639 87 3 gnns gnn NOUN bracis-33639 87 4 have have AUX bracis-33639 87 5 shown show VERB bracis-33639 87 6 substantial substantial ADJ bracis-33639 87 7 advancements advancement NOUN bracis-33639 87 8 in in ADP bracis-33639 87 9 various various ADJ bracis-33639 87 10 applications application NOUN bracis-33639 87 11 , , PUNCT bracis-33639 87 12 such such ADJ bracis-33639 87 13 as as ADP bracis-33639 87 14 traffic traffic NOUN bracis-33639 87 15 prediction prediction NOUN bracis-33639 87 16 , , PUNCT bracis-33639 87 17 social social ADJ bracis-33639 87 18 network network NOUN bracis-33639 87 19 analysis analysis NOUN bracis-33639 87 20 , , PUNCT bracis-33639 87 21 and and CCONJ bracis-33639 87 22 financial financial ADJ bracis-33639 87 23 forecasting forecasting NOUN bracis-33639 87 24 , , PUNCT bracis-33639 87 25 effectively effectively ADV bracis-33639 87 26 modeling model VERB bracis-33639 87 27 the the DET bracis-33639 87 28 temporal temporal ADJ bracis-33639 87 29 evolution evolution NOUN bracis-33639 87 30 of of ADP bracis-33639 87 31 graph graph NOUN bracis-33639 87 32 data datum NOUN bracis-33639 87 33 . . PUNCT bracis-33639 88 1 in in ADP bracis-33639 88 2 the the DET bracis-33639 88 3 context context NOUN bracis-33639 88 4 of of ADP bracis-33639 88 5 crimes crime NOUN bracis-33639 88 6 , , PUNCT bracis-33639 88 7 han han PROPN bracis-33639 88 8 et et PROPN bracis-33639 88 9 al al PROPN bracis-33639 88 10 . . PUNCT bracis-33639 89 1 [ [ X bracis-33639 89 2 10 10 NUM bracis-33639 89 3 ] ] PUNCT bracis-33639 89 4 developed develop VERB bracis-33639 89 5 a a DET bracis-33639 89 6 crime crime NOUN bracis-33639 89 7 prediction prediction NOUN bracis-33639 89 8 framework framework NOUN bracis-33639 89 9 using use VERB bracis-33639 89 10 graph graph NOUN bracis-33639 89 11 convolutional convolutional ADJ bracis-33639 89 12 networks network NOUN bracis-33639 89 13 to to PART bracis-33639 89 14 model model VERB bracis-33639 89 15 the the DET bracis-33639 89 16 spatial spatial ADJ bracis-33639 89 17 dependencies dependency NOUN bracis-33639 89 18 between between ADP bracis-33639 89 19 different different ADJ bracis-33639 89 20 regions region NOUN bracis-33639 89 21 and and CCONJ bracis-33639 89 22 long long ADJ bracis-33639 89 23 short short ADJ bracis-33639 89 24 - - PUNCT bracis-33639 89 25 term term NOUN bracis-33639 89 26 memory memory NOUN bracis-33639 89 27 ( ( PUNCT bracis-33639 89 28 lstm lstm NOUN bracis-33639 89 29 ) ) PUNCT bracis-33639 89 30 networks network NOUN bracis-33639 89 31 to to PART bracis-33639 89 32 capture capture VERB bracis-33639 89 33 temporal temporal ADJ bracis-33639 89 34 patterns pattern NOUN bracis-33639 89 35 . . PUNCT bracis-33639 90 1 jin jin NOUN bracis-33639 90 2 et et PROPN bracis-33639 90 3 al al PROPN bracis-33639 90 4 . . PUNCT bracis-33639 90 5     SPACE bracis-33639 91 1 [ [ X bracis-33639 91 2 13 13 NUM bracis-33639 91 3 ] ] PUNCT bracis-33639 91 4 introduced introduce VERB bracis-33639 91 5 a a DET bracis-33639 91 6 spatio spatio NOUN bracis-33639 91 7 - - PUNCT bracis-33639 91 8 temporal temporal ADJ bracis-33639 91 9 graph graph NOUN bracis-33639 91 10 convolutional convolutional ADJ bracis-33639 91 11 network network NOUN bracis-33639 91 12 ( ( PUNCT bracis-33639 91 13 st st NOUN bracis-33639 91 14 - - PUNCT bracis-33639 91 15 gcn gcn NOUN bracis-33639 91 16 ) ) PUNCT bracis-33639 91 17 that that PRON bracis-33639 91 18 combines combine VERB bracis-33639 91 19 gcns gcns PROPN bracis-33639 91 20 with with ADP bracis-33639 91 21 recurrent recurrent ADJ bracis-33639 91 22 neural neural ADJ bracis-33639 91 23 networks network NOUN bracis-33639 91 24 to to PART bracis-33639 91 25 predict predict VERB bracis-33639 91 26 crime crime NOUN bracis-33639 91 27 hotspots hotspot NOUN bracis-33639 91 28 , , PUNCT bracis-33639 91 29 learning learn VERB bracis-33639 91 30 both both CCONJ bracis-33639 91 31 spatial spatial ADJ bracis-33639 91 32 and and CCONJ bracis-33639 91 33 temporal temporal ADJ bracis-33639 91 34 features feature NOUN bracis-33639 91 35 from from ADP bracis-33639 91 36 crime crime NOUN bracis-33639 91 37 data datum NOUN bracis-33639 91 38 . . PUNCT bracis-33639 92 1 the the DET bracis-33639 92 2 discussion discussion NOUN bracis-33639 92 3 above above ADV bracis-33639 92 4 has have VERB bracis-33639 92 5 only only ADV bracis-33639 92 6 the the DET bracis-33639 92 7 intent intent NOUN bracis-33639 92 8 of of ADP bracis-33639 92 9 contextualizing contextualize VERB bracis-33639 92 10 our our PRON bracis-33639 92 11 contribution contribution NOUN bracis-33639 92 12 . . PUNCT bracis-33639 93 1 a a DET bracis-33639 93 2 more more ADV bracis-33639 93 3 comprehensive comprehensive ADJ bracis-33639 93 4 discussion discussion NOUN bracis-33639 93 5 about about ADP bracis-33639 93 6 the the DET bracis-33639 93 7 use use NOUN bracis-33639 93 8 of of ADP bracis-33639 93 9 machine machine NOUN bracis-33639 93 10 and and CCONJ bracis-33639 93 11 deep deep ADJ bracis-33639 93 12 learning learning NOUN bracis-33639 93 13 for for ADP bracis-33639 93 14 crime crime NOUN bracis-33639 93 15 forecasting forecasting NOUN bracis-33639 93 16 can can AUX bracis-33639 93 17 be be AUX bracis-33639 93 18 found find VERB bracis-33639 93 19 in in ADP bracis-33639 93 20 several several ADJ bracis-33639 93 21 surveys survey NOUN bracis-33639 93 22 bout bout VERB bracis-33639 93 23 the the DET bracis-33639 93 24 theme theme NOUN bracis-33639 93 25     SPACE bracis-33639 94 1 [ [ X bracis-33639 94 2 5 5 NUM bracis-33639 94 3 , , PUNCT bracis-33639 94 4 18 18 NUM bracis-33639 94 5 , , PUNCT bracis-33639 94 6 29 29 NUM bracis-33639 94 7 ] ] PUNCT bracis-33639 94 8 . . PUNCT bracis-33639 94 9 3 3 NUM bracis-33639 94 10 data datum NOUN bracis-33639 94 11 modeling modeling NOUN bracis-33639 94 12 for for ADP bracis-33639 94 13 this this DET bracis-33639 94 14 study study NOUN bracis-33639 94 15 , , PUNCT bracis-33639 94 16 we we PRON bracis-33639 94 17 utilized utilize VERB bracis-33639 94 18 crime crime NOUN bracis-33639 94 19 data datum NOUN bracis-33639 94 20 from from ADP bracis-33639 94 21 the the DET bracis-33639 94 22 são são PROPN bracis-33639 94 23 paulo paulo PROPN bracis-33639 94 24 department department PROPN bracis-33639 94 25 of of ADP bracis-33639 94 26 public public ADJ bracis-33639 94 27 safety safety NOUN bracis-33639 95 1 [ [ X bracis-33639 95 2 27 27 NUM bracis-33639 95 3 ] ] PUNCT bracis-33639 95 4 , , PUNCT bracis-33639 95 5 which which PRON bracis-33639 95 6 includes include VERB bracis-33639 95 7 detailed detailed ADJ bracis-33639 95 8 information information NOUN bracis-33639 95 9 on on ADP bracis-33639 95 10 crime crime NOUN bracis-33639 95 11 occurrences occurrence NOUN bracis-33639 95 12 with with ADP bracis-33639 95 13 latitude latitude NOUN bracis-33639 95 14 and and CCONJ bracis-33639 95 15 longitude longitude ADJ bracis-33639 95 16 coordinates coordinate NOUN bracis-33639 95 17 . . PUNCT bracis-33639 96 1 to to PART bracis-33639 96 2 complement complement VERB bracis-33639 96 3 this this DET bracis-33639 96 4 data datum NOUN bracis-33639 96 5 , , PUNCT bracis-33639 96 6 we we PRON bracis-33639 96 7 extracted extract VERB bracis-33639 96 8 street street NOUN bracis-33639 96 9 network network NOUN bracis-33639 96 10 data datum NOUN bracis-33639 96 11 using use VERB bracis-33639 96 12 the the DET bracis-33639 96 13 osmnx osmnx PROPN bracis-33639 96 14 python python NOUN bracis-33639 96 15 package package NOUN bracis-33639 96 16 [ [ X bracis-33639 96 17 3 3 NUM bracis-33639 96 18 ] ] PUNCT bracis-33639 96 19 . . PUNCT bracis-33639 97 1 the the DET bracis-33639 97 2 street street NOUN bracis-33639 97 3 network network NOUN bracis-33639 97 4 data datum NOUN bracis-33639 97 5 is be AUX bracis-33639 97 6 used use VERB bracis-33639 97 7 to to PART bracis-33639 97 8 construct construct VERB bracis-33639 97 9 graphs graph NOUN bracis-33639 97 10 representing represent VERB bracis-33639 97 11 streets street NOUN bracis-33639 97 12 as as ADP bracis-33639 97 13 nodes node NOUN bracis-33639 97 14 and and CCONJ bracis-33639 97 15 their their PRON bracis-33639 97 16 connections connection NOUN bracis-33639 97 17 as as ADP bracis-33639 97 18 edges edge NOUN bracis-33639 97 19 . . PUNCT bracis-33639 98 1 each each DET bracis-33639 98 2 node node NOUN bracis-33639 98 3 is be AUX bracis-33639 98 4 labeled label VERB bracis-33639 98 5 to to PART bracis-33639 98 6 indicate indicate VERB bracis-33639 98 7 the the DET bracis-33639 98 8 presence presence NOUN bracis-33639 98 9 or or CCONJ bracis-33639 98 10 absence absence NOUN bracis-33639 98 11 of of ADP bracis-33639 98 12 a a DET bracis-33639 98 13 crime crime NOUN bracis-33639 98 14 , , PUNCT bracis-33639 98 15 making make VERB bracis-33639 98 16 the the DET bracis-33639 98 17 labels label NOUN bracis-33639 98 18 dynamic dynamic ADJ bracis-33639 98 19 as as SCONJ bracis-33639 98 20 they they PRON bracis-33639 98 21 evolve evolve VERB bracis-33639 98 22 over over ADP bracis-33639 98 23 time time NOUN bracis-33639 98 24 . . PUNCT bracis-33639 99 1 the the DET bracis-33639 99 2 crime crime NOUN bracis-33639 99 3 and and CCONJ bracis-33639 99 4 street street NOUN bracis-33639 99 5 network network NOUN bracis-33639 99 6 data datum NOUN bracis-33639 99 7 are be AUX bracis-33639 99 8 integrated integrate VERB bracis-33639 99 9 to to PART bracis-33639 99 10 form form VERB bracis-33639 99 11 a a DET bracis-33639 99 12 spatiotemporal spatiotemporal ADJ bracis-33639 99 13 dataset dataset NOUN bracis-33639 99 14 . . PUNCT bracis-33639 100 1 this this DET bracis-33639 100 2 integration integration NOUN bracis-33639 100 3 is be AUX bracis-33639 100 4 crucial crucial ADJ bracis-33639 100 5 for for ADP bracis-33639 100 6 creating create VERB bracis-33639 100 7 the the DET bracis-33639 100 8 dynamic dynamic ADJ bracis-33639 100 9 graph graph NOUN bracis-33639 100 10 structures structure NOUN bracis-33639 100 11 necessary necessary ADJ bracis-33639 100 12 for for ADP bracis-33639 100 13 our our PRON bracis-33639 100 14 models model NOUN bracis-33639 100 15 . . PUNCT bracis-33639 101 1 the the DET bracis-33639 101 2 resulting result VERB bracis-33639 101 3 dataset dataset NOUN bracis-33639 101 4 consists consist NOUN bracis-33639 101 5 of of ADP bracis-33639 101 6 monthly monthly ADJ bracis-33639 101 7 snapshots snapshot NOUN bracis-33639 101 8 of of ADP bracis-33639 101 9 the the DET bracis-33639 101 10 graph graph NOUN bracis-33639 101 11 spanning span VERB bracis-33639 101 12 two two NUM bracis-33639 101 13 years year NOUN bracis-33639 101 14 , , PUNCT bracis-33639 101 15 resulting result VERB bracis-33639 101 16 in in ADP bracis-33639 101 17 24 24 NUM bracis-33639 101 18 graphs graph NOUN bracis-33639 101 19 in in ADP bracis-33639 101 20 total total NOUN bracis-33639 101 21 . . PUNCT bracis-33639 102 1 in in ADP bracis-33639 102 2 this this DET bracis-33639 102 3 work work NOUN bracis-33639 102 4 , , PUNCT bracis-33639 102 5 we we PRON bracis-33639 102 6 have have AUX bracis-33639 102 7 used use VERB bracis-33639 102 8 months month NOUN bracis-33639 102 9 as as ADP bracis-33639 102 10 a a DET bracis-33639 102 11 time time NOUN bracis-33639 102 12 duration duration NOUN bracis-33639 102 13 , , PUNCT bracis-33639 102 14 but but CCONJ bracis-33639 102 15 our our PRON bracis-33639 102 16 data data NOUN bracis-33639 102 17 modeling model VERB bracis-33639 102 18 pipeline pipeline NOUN bracis-33639 102 19 can can AUX bracis-33639 102 20 also also ADV bracis-33639 102 21 be be AUX bracis-33639 102 22 used use VERB bracis-33639 102 23 for for ADP bracis-33639 102 24 days day NOUN bracis-33639 102 25 and and CCONJ bracis-33639 102 26 years year NOUN bracis-33639 102 27 , , PUNCT bracis-33639 102 28 depending depend VERB bracis-33639 102 29 on on ADP bracis-33639 102 30 the the DET bracis-33639 102 31 analysis analysis NOUN bracis-33639 102 32 . . PUNCT bracis-33639 103 1 we we PRON bracis-33639 103 2 decided decide VERB bracis-33639 103 3 to to PART bracis-33639 103 4 use use VERB bracis-33639 103 5 monthly monthly ADJ bracis-33639 103 6 intervals interval NOUN bracis-33639 103 7 as as ADP bracis-33639 103 8 a a DET bracis-33639 103 9 temporal temporal ADJ bracis-33639 103 10 range range NOUN bracis-33639 103 11 to to PART bracis-33639 103 12 reduce reduce VERB bracis-33639 103 13 the the DET bracis-33639 103 14 sparsity sparsity NOUN bracis-33639 103 15 effect effect NOUN bracis-33639 103 16 in in ADP bracis-33639 103 17 the the DET bracis-33639 103 18 data datum NOUN bracis-33639 103 19 . . PUNCT bracis-33639 104 1 if if SCONJ bracis-33639 104 2 processed process VERB bracis-33639 104 3 daily daily ADV bracis-33639 104 4 , , PUNCT bracis-33639 104 5 the the DET bracis-33639 104 6 ratio ratio NOUN bracis-33639 104 7 of of ADP bracis-33639 104 8 occurrences occurrence NOUN bracis-33639 104 9 to to ADP bracis-33639 104 10 non non NOUN bracis-33639 104 11 - - NOUN bracis-33639 104 12 occurrences occurrence NOUN bracis-33639 104 13 would would AUX bracis-33639 104 14 be be AUX bracis-33639 104 15 around around ADP bracis-33639 104 16 1/100 1/100 NUM bracis-33639 104 17 , , PUNCT bracis-33639 104 18 whereas whereas SCONJ bracis-33639 104 19 with with ADP bracis-33639 104 20 monthly monthly ADJ bracis-33639 104 21 processing processing NOUN bracis-33639 104 22 , , PUNCT bracis-33639 104 23 it it PRON bracis-33639 104 24 is be AUX bracis-33639 104 25 around around ADP bracis-33639 104 26 1/5 1/5 NUM bracis-33639 104 27 . . PUNCT bracis-33639 105 1 in in ADP bracis-33639 105 2 more more ADJ bracis-33639 105 3 details detail NOUN bracis-33639 105 4 , , PUNCT bracis-33639 105 5 the the DET bracis-33639 105 6 graph graph NOUN bracis-33639 105 7 construction construction NOUN bracis-33639 105 8 process process NOUN bracis-33639 105 9 involved involve VERB bracis-33639 105 10 the the DET bracis-33639 105 11 following follow VERB bracis-33639 105 12 steps step NOUN bracis-33639 105 13 : : PUNCT bracis-33639 105 14 1 1 X bracis-33639 105 15 . . X bracis-33639 105 16 downloading download VERB bracis-33639 105 17 street street NOUN bracis-33639 105 18 graph graph NOUN bracis-33639 105 19 from from ADP bracis-33639 105 20 given give VERB bracis-33639 105 21 region region NOUN bracis-33639 105 22 : : PUNCT bracis-33639 105 23 we we PRON bracis-33639 105 24 use use VERB bracis-33639 105 25 osmnx osmnx PROPN bracis-33639 105 26 [ [ X bracis-33639 105 27 3 3 NUM bracis-33639 105 28 ] ] PUNCT bracis-33639 105 29 to to PART bracis-33639 105 30 download download VERB bracis-33639 105 31 street street NOUN bracis-33639 105 32 graphs graph NOUN bracis-33639 105 33 from from ADP bracis-33639 105 34 openstreetmaps openstreetmap NOUN bracis-33639 105 35 . . PUNCT bracis-33639 106 1 our our PRON bracis-33639 106 2 code code NOUN bracis-33639 106 3 allows allow VERB bracis-33639 106 4 us we PRON bracis-33639 106 5 to to PART bracis-33639 106 6 input input VERB bracis-33639 106 7 two two NUM bracis-33639 106 8 types type NOUN bracis-33639 106 9 of of ADP bracis-33639 106 10 parameters parameter NOUN bracis-33639 106 11 at at ADP bracis-33639 106 12 this this DET bracis-33639 106 13 stage stage NOUN bracis-33639 106 14 : : PUNCT bracis-33639 106 15 a a DET bracis-33639 106 16 string string NOUN bracis-33639 106 17 containing contain VERB bracis-33639 106 18 the the DET bracis-33639 106 19 location location NOUN bracis-33639 106 20 from from ADP bracis-33639 106 21 which which PRON bracis-33639 106 22 we we PRON bracis-33639 106 23 want want VERB bracis-33639 106 24 to to PART bracis-33639 106 25 fetch fetch VERB bracis-33639 106 26 the the DET bracis-33639 106 27 graph graph NOUN bracis-33639 106 28 , , PUNCT bracis-33639 106 29 or or CCONJ bracis-33639 106 30 a a DET bracis-33639 106 31 polygon polygon NOUN bracis-33639 106 32 formed form VERB bracis-33639 106 33 by by ADP bracis-33639 106 34 a a DET bracis-33639 106 35 list list NOUN bracis-33639 106 36 of of ADP bracis-33639 106 37 latitude latitude NOUN bracis-33639 106 38 / / SYM bracis-33639 106 39 longitude longitude NOUN bracis-33639 106 40 coordinates coordinate NOUN bracis-33639 106 41 defining define VERB bracis-33639 106 42 the the DET bracis-33639 106 43 region region NOUN bracis-33639 106 44 of of ADP bracis-33639 106 45 interest interest NOUN bracis-33639 106 46 . . PUNCT bracis-33639 107 1 in in ADP bracis-33639 107 2 this this DET bracis-33639 107 3 study study NOUN bracis-33639 107 4 , , PUNCT bracis-33639 107 5 we we PRON bracis-33639 107 6 restricted restrict VERB bracis-33639 107 7 our our PRON bracis-33639 107 8 graph graph NOUN bracis-33639 107 9 to to ADP bracis-33639 107 10 the the DET bracis-33639 107 11 central central ADJ bracis-33639 107 12 region region NOUN bracis-33639 107 13 of of ADP bracis-33639 107 14 são são PROPN bracis-33639 107 15 paulo paulo PROPN bracis-33639 107 16 . . PUNCT bracis-33639 108 1 2 2 X bracis-33639 108 2 . . X bracis-33639 108 3 processing processing NOUN bracis-33639 108 4 street street NOUN bracis-33639 108 5 graph graph NOUN bracis-33639 108 6 : : PUNCT bracis-33639 108 7 in in ADP bracis-33639 108 8 step step NOUN bracis-33639 108 9 1 1 NUM bracis-33639 108 10 , , PUNCT bracis-33639 108 11 we we PRON bracis-33639 108 12 acquire acquire VERB bracis-33639 108 13 a a DET bracis-33639 108 14 directed direct VERB bracis-33639 108 15 graph graph NOUN bracis-33639 108 16 where where SCONJ bracis-33639 108 17 streets street NOUN bracis-33639 108 18 are be AUX bracis-33639 108 19 depicted depict VERB bracis-33639 108 20 as as ADP bracis-33639 108 21 edges edge NOUN bracis-33639 108 22 and and CCONJ bracis-33639 108 23 intersections intersection NOUN bracis-33639 108 24 as as ADP bracis-33639 108 25 vertices vertex NOUN bracis-33639 108 26 , , PUNCT bracis-33639 108 27 encompassing encompass VERB bracis-33639 108 28 diverse diverse ADJ bracis-33639 108 29 transportation transportation NOUN bracis-33639 108 30 networks network NOUN bracis-33639 108 31 . . PUNCT bracis-33639 109 1 in in ADP bracis-33639 109 2 this this DET bracis-33639 109 3 step step NOUN bracis-33639 109 4 , , PUNCT bracis-33639 109 5 we we PRON bracis-33639 109 6 unify unify VERB bracis-33639 109 7 all all DET bracis-33639 109 8 types type NOUN bracis-33639 109 9 of of ADP bracis-33639 109 10 transportation transportation NOUN bracis-33639 109 11 networks network NOUN bracis-33639 109 12 and and CCONJ bracis-33639 109 13 invert invert VERB bracis-33639 109 14 the the DET bracis-33639 109 15 roles role NOUN bracis-33639 109 16 of of ADP bracis-33639 109 17 vertices vertex NOUN bracis-33639 109 18 and and CCONJ bracis-33639 109 19 edges edge NOUN bracis-33639 109 20 : : PUNCT bracis-33639 109 21 street street NOUN bracis-33639 109 22 segments segment NOUN bracis-33639 109 23 are be AUX bracis-33639 109 24 redefined redefine VERB bracis-33639 109 25 as as ADP bracis-33639 109 26 vertices vertex NOUN bracis-33639 109 27 , , PUNCT bracis-33639 109 28 while while SCONJ bracis-33639 109 29 intersections intersection NOUN bracis-33639 109 30 become become VERB bracis-33639 109 31 edges edge NOUN bracis-33639 109 32 , , PUNCT bracis-33639 109 33 resulting result VERB bracis-33639 109 34 in in ADP bracis-33639 109 35 a a DET bracis-33639 109 36 undirected undirected ADJ bracis-33639 109 37 graph graph NOUN bracis-33639 109 38 . . PUNCT bracis-33639 110 1 3 3 X bracis-33639 110 2 . . X bracis-33639 110 3 extracting extract VERB bracis-33639 110 4 spatial spatial ADJ bracis-33639 110 5 features feature NOUN bracis-33639 110 6 : : PUNCT bracis-33639 110 7 for for ADP bracis-33639 110 8 each each DET bracis-33639 110 9 vertex vertex NOUN bracis-33639 110 10 , , PUNCT bracis-33639 110 11 we we PRON bracis-33639 110 12 associate associate VERB bracis-33639 110 13 a a DET bracis-33639 110 14 feature feature NOUN bracis-33639 110 15 vector vector NOUN bracis-33639 110 16 containing contain VERB bracis-33639 110 17 spatial spatial ADJ bracis-33639 110 18 information information NOUN bracis-33639 110 19 . . PUNCT bracis-33639 111 1 these these DET bracis-33639 111 2 features feature NOUN bracis-33639 111 3 are be AUX bracis-33639 111 4 static static ADJ bracis-33639 111 5 and and CCONJ bracis-33639 111 6 sourced source VERB bracis-33639 111 7 from from ADP bracis-33639 111 8 the the DET bracis-33639 111 9 public public ADJ bracis-33639 111 10 geosampa geosampa NOUN bracis-33639 111 11 website website NOUN bracis-33639 111 12 [ [ X bracis-33639 111 13 24 24 NUM bracis-33639 111 14 ] ] PUNCT bracis-33639 111 15 , , PUNCT bracis-33639 111 16 which which PRON bracis-33639 111 17 provides provide VERB bracis-33639 111 18 tools tool NOUN bracis-33639 111 19 for for ADP bracis-33639 111 20 extracting extract VERB bracis-33639 111 21 various various ADJ bracis-33639 111 22 data datum NOUN bracis-33639 111 23 about about ADP bracis-33639 111 24 the the DET bracis-33639 111 25 city city NOUN bracis-33639 111 26 . . PUNCT bracis-33639 112 1 the the DET bracis-33639 112 2 features feature NOUN bracis-33639 112 3 selected select VERB bracis-33639 112 4 for for ADP bracis-33639 112 5 testing testing NOUN bracis-33639 112 6 include include VERB bracis-33639 112 7 points point NOUN bracis-33639 112 8 of of ADP bracis-33639 112 9 interest interest NOUN bracis-33639 112 10 , , PUNCT bracis-33639 112 11 such such ADJ bracis-33639 112 12 as as ADP bracis-33639 112 13 the the DET bracis-33639 112 14 number number NOUN bracis-33639 112 15 of of ADP bracis-33639 112 16 health health NOUN bracis-33639 112 17 , , PUNCT bracis-33639 112 18 cultural cultural ADJ bracis-33639 112 19 , , PUNCT bracis-33639 112 20 security security NOUN bracis-33639 112 21 , , PUNCT bracis-33639 112 22 education education NOUN bracis-33639 112 23 , , PUNCT bracis-33639 112 24 and and CCONJ bracis-33639 112 25 social social ADJ bracis-33639 112 26 assistance assistance NOUN bracis-33639 112 27 services service NOUN bracis-33639 112 28 located locate VERB bracis-33639 112 29 within within ADP bracis-33639 112 30 a a DET bracis-33639 112 31 predetermined predetermine VERB bracis-33639 112 32 distance distance NOUN bracis-33639 112 33 from from ADP bracis-33639 112 34 a a DET bracis-33639 112 35 node node NOUN bracis-33639 112 36 where where SCONJ bracis-33639 112 37 a a DET bracis-33639 112 38 crime crime NOUN bracis-33639 112 39 occurred occur VERB bracis-33639 112 40 . . PUNCT bracis-33639 113 1 4 4 X bracis-33639 113 2 . . X bracis-33639 113 3 projecting project VERB bracis-33639 113 4 crime crime NOUN bracis-33639 113 5 data datum NOUN bracis-33639 113 6 : : PUNCT bracis-33639 113 7 crime crime NOUN bracis-33639 113 8 data datum NOUN bracis-33639 113 9 is be AUX bracis-33639 113 10 extracted extract VERB bracis-33639 113 11 from from ADP bracis-33639 113 12 são são PROPN bracis-33639 113 13 paulo paulo PROPN bracis-33639 113 14 ’s ’s PART bracis-33639 113 15 department department PROPN bracis-33639 113 16 of of ADP bracis-33639 113 17 public public ADJ bracis-33639 113 18 safety safety NOUN bracis-33639 113 19 [ [ X bracis-33639 113 20 27 27 NUM bracis-33639 113 21 ] ] PUNCT bracis-33639 113 22 . . PUNCT bracis-33639 114 1 each each DET bracis-33639 114 2 occurrence occurrence NOUN bracis-33639 114 3 is be AUX bracis-33639 114 4 associated associate VERB bracis-33639 114 5 with with ADP bracis-33639 114 6 a a DET bracis-33639 114 7 latitude latitude NOUN bracis-33639 114 8 and and CCONJ bracis-33639 114 9 longitude longitude ADJ bracis-33639 114 10 coordinate coordinate NOUN bracis-33639 114 11 , , PUNCT bracis-33639 114 12 which which PRON bracis-33639 114 13 is be AUX bracis-33639 114 14 then then ADV bracis-33639 114 15 projected project VERB bracis-33639 114 16 onto onto ADP bracis-33639 114 17 the the DET bracis-33639 114 18 respective respective ADJ bracis-33639 114 19 street street NOUN bracis-33639 114 20 vertex vertex NOUN bracis-33639 114 21 . . PUNCT bracis-33639 115 1 crimes crime NOUN bracis-33639 115 2 with with ADP bracis-33639 115 3 coordinates coordinate NOUN bracis-33639 115 4 outside outside ADP bracis-33639 115 5 the the DET bracis-33639 115 6 region region NOUN bracis-33639 115 7 define define VERB bracis-33639 115 8 in in ADP bracis-33639 115 9 step step NOUN bracis-33639 115 10 1 1 NUM bracis-33639 115 11 are be AUX bracis-33639 115 12 removed remove VERB bracis-33639 115 13 . . PUNCT bracis-33639 116 1 the the DET bracis-33639 116 2 complete complete ADJ bracis-33639 116 3 data datum NOUN bracis-33639 116 4 processing processing NOUN bracis-33639 116 5 pipeline pipeline NOUN bracis-33639 116 6 can can AUX bracis-33639 116 7 be be AUX bracis-33639 116 8 found find VERB bracis-33639 116 9 in in ADP bracis-33639 116 10 our our PRON bracis-33639 116 11 github github NOUN bracis-33639 116 12 repositoryfootnote repositoryfootnote VERB bracis-33639 116 13 1 1 NUM bracis-33639 116 14 . . PUNCT bracis-33639 117 1 the the DET bracis-33639 117 2 final final ADJ bracis-33639 117 3 result result NOUN bracis-33639 117 4 of of ADP bracis-33639 117 5 our our PRON bracis-33639 117 6 data data NOUN bracis-33639 117 7 processing processing NOUN bracis-33639 117 8 pipeline pipeline NOUN bracis-33639 117 9 is be AUX bracis-33639 117 10 illustrated illustrate VERB bracis-33639 117 11 in in ADP bracis-33639 117 12 fig fig NOUN bracis-33639 117 13 . . PUNCT bracis-33639 117 14     SPACE bracis-33639 118 1 1 1 NUM bracis-33639 118 2 . . X bracis-33639 118 3 figure figure NOUN bracis-33639 118 4     SPACE bracis-33639 118 5 1 1 NUM bracis-33639 118 6 represents represent VERB bracis-33639 118 7 data datum NOUN bracis-33639 118 8 from from ADP bracis-33639 118 9 the the DET bracis-33639 118 10 month month NOUN bracis-33639 118 11 immediately immediately ADV bracis-33639 118 12 following follow VERB bracis-33639 118 13 the the DET bracis-33639 118 14 training training NOUN bracis-33639 118 15 period period NOUN bracis-33639 118 16 . . PUNCT bracis-33639 119 1 the the DET bracis-33639 119 2 street street NOUN bracis-33639 119 3 segments segment NOUN bracis-33639 119 4 with with ADP bracis-33639 119 5 green green ADJ bracis-33639 119 6 lines line NOUN bracis-33639 119 7 represent represent VERB bracis-33639 119 8 the the DET bracis-33639 119 9 non non ADJ bracis-33639 119 10 - - ADJ bracis-33639 119 11 crime crime ADJ bracis-33639 119 12 nodes node NOUN bracis-33639 119 13 ( ( PUNCT bracis-33639 119 14 streets street NOUN bracis-33639 119 15 ) ) PUNCT bracis-33639 119 16 , , PUNCT bracis-33639 119 17 and and CCONJ bracis-33639 119 18 the the DET bracis-33639 119 19 red red ADJ bracis-33639 119 20 lines line NOUN bracis-33639 119 21 represent represent VERB bracis-33639 119 22 crime crime NOUN bracis-33639 119 23 nodes node NOUN bracis-33639 119 24 . . PUNCT bracis-33639 120 1 fig fig NOUN bracis-33639 120 2 . . PUNCT bracis-33639 121 1 1 1 NUM bracis-33639 121 2 . . X bracis-33639 121 3 street street NOUN bracis-33639 121 4 map map NOUN bracis-33639 121 5 graph graph NOUN bracis-33639 121 6 for for ADP bracis-33639 121 7 the the DET bracis-33639 121 8 central central ADJ bracis-33639 121 9 region region NOUN bracis-33639 121 10 of of ADP bracis-33639 121 11 são são PROPN bracis-33639 121 12 paulo paulo PROPN bracis-33639 121 13 . . PUNCT bracis-33639 122 1 crime crime NOUN bracis-33639 122 2 nodes node NOUN bracis-33639 122 3 are be AUX bracis-33639 122 4 marked mark VERB bracis-33639 122 5 in in ADP bracis-33639 122 6 red red ADJ bracis-33639 122 7 , , PUNCT bracis-33639 122 8 and and CCONJ bracis-33639 122 9 non non ADJ bracis-33639 122 10 - - ADJ bracis-33639 122 11 crime crime ADJ bracis-33639 122 12 nodes node NOUN bracis-33639 122 13 are be AUX bracis-33639 122 14 marked mark VERB bracis-33639 122 15 in in ADP bracis-33639 122 16 green green ADJ bracis-33639 122 17 . . PUNCT bracis-33639 123 1 ( ( PUNCT bracis-33639 123 2 color color NOUN bracis-33639 123 3 figure figure NOUN bracis-33639 123 4 online online ADV bracis-33639 123 5 ) ) PUNCT bracis-33639 123 6 full full ADJ bracis-33639 123 7 size size NOUN bracis-33639 123 8 image image NOUN bracis-33639 123 9 4 4 NUM bracis-33639 123 10 models model NOUN bracis-33639 123 11 this this DET bracis-33639 123 12 paper paper NOUN bracis-33639 123 13 utilizes utilize VERB bracis-33639 123 14 two two NUM bracis-33639 123 15 distinct distinct ADJ bracis-33639 123 16 graph graph NOUN bracis-33639 123 17 neural neural ADJ bracis-33639 123 18 networks network NOUN bracis-33639 123 19 ( ( PUNCT bracis-33639 123 20 gnns gnns NOUN bracis-33639 123 21 ) ) PUNCT bracis-33639 123 22 architectures architecture NOUN bracis-33639 123 23 for for ADP bracis-33639 123 24 crime crime NOUN bracis-33639 123 25 prediction prediction NOUN bracis-33639 123 26 : : PUNCT bracis-33639 123 27 dysat dysat VERB bracis-33639 123 28 [ [ X bracis-33639 123 29 26 26 NUM bracis-33639 123 30 ] ] PUNCT bracis-33639 123 31 and and CCONJ bracis-33639 123 32 evolvegcn evolvegcn PROPN bracis-33639 124 1 [ [ X bracis-33639 124 2 22 22 NUM bracis-33639 124 3 ] ] PUNCT bracis-33639 124 4 . . PUNCT bracis-33639 125 1 the the DET bracis-33639 125 2 choice choice NOUN bracis-33639 125 3 of of ADP bracis-33639 125 4 dysat dysat NOUN bracis-33639 125 5 and and CCONJ bracis-33639 125 6 evolvegcn evolvegcn PROPN bracis-33639 125 7 is be AUX bracis-33639 125 8 motivated motivate VERB bracis-33639 125 9 by by ADP bracis-33639 125 10 their their PRON bracis-33639 125 11 ability ability NOUN bracis-33639 125 12 to to PART bracis-33639 125 13 handle handle VERB bracis-33639 125 14 dynamic dynamic ADJ bracis-33639 125 15 graphs graph NOUN bracis-33639 125 16 effectively effectively ADV bracis-33639 125 17 [ [ X bracis-33639 125 18 16 16 NUM bracis-33639 125 19 ] ] PUNCT bracis-33639 125 20 , , PUNCT bracis-33639 125 21 making make VERB bracis-33639 125 22 them they PRON bracis-33639 125 23 eligible eligible ADJ bracis-33639 125 24 for for ADP bracis-33639 125 25 our our PRON bracis-33639 125 26 analysis analysis NOUN bracis-33639 125 27 . . PUNCT bracis-33639 126 1 4.1 4.1 NUM bracis-33639 126 2 dysat dysat ADJ bracis-33639 126 3 dynamic dynamic ADJ bracis-33639 126 4 self self NOUN bracis-33639 126 5 - - PUNCT bracis-33639 126 6 attention attention NOUN bracis-33639 126 7 network network NOUN bracis-33639 126 8 ( ( PUNCT bracis-33639 126 9 dysat dysat NOUN bracis-33639 126 10 ) ) PUNCT bracis-33639 126 11 [ [ X bracis-33639 126 12 26 26 NUM bracis-33639 126 13 ] ] PUNCT bracis-33639 126 14 is be AUX bracis-33639 126 15 an an DET bracis-33639 126 16 unsupervised unsupervised ADJ bracis-33639 126 17 graph graph NOUN bracis-33639 126 18 embedding embed VERB bracis-33639 126 19 model model NOUN bracis-33639 126 20 that that PRON bracis-33639 126 21 learns learn VERB bracis-33639 126 22 a a DET bracis-33639 126 23 latent latent NOUN bracis-33639 126 24 node node NOUN bracis-33639 126 25 representation representation NOUN bracis-33639 126 26 to to PART bracis-33639 126 27 capture capture VERB bracis-33639 126 28 dynamic dynamic ADJ bracis-33639 126 29 graph graph NOUN bracis-33639 126 30 representation representation NOUN bracis-33639 126 31 . . PUNCT bracis-33639 127 1 it it PRON bracis-33639 127 2 computes compute VERB bracis-33639 127 3 node node ADJ bracis-33639 127 4 representations representation NOUN bracis-33639 127 5 through through ADP bracis-33639 127 6 self self NOUN bracis-33639 127 7 - - PUNCT bracis-33639 127 8 attention attention NOUN bracis-33639 127 9 along along ADP bracis-33639 127 10 two two NUM bracis-33639 127 11 dimensions dimension NOUN bracis-33639 127 12 ( ( PUNCT bracis-33639 127 13 layers layer NOUN bracis-33639 127 14 ): ): PUNCT bracis-33639 127 15 structural structural ADJ bracis-33639 127 16 neighborhood neighborhood NOUN bracis-33639 127 17 and and CCONJ bracis-33639 127 18 temporal temporal ADJ bracis-33639 127 19 dynamics dynamic NOUN bracis-33639 127 20 . . PUNCT bracis-33639 128 1 the the DET bracis-33639 128 2 structural structural ADJ bracis-33639 128 3 attention attention NOUN bracis-33639 128 4 layer layer NOUN bracis-33639 128 5 extracts extract NOUN bracis-33639 128 6 features feature NOUN bracis-33639 128 7 from from ADP bracis-33639 128 8 local local ADJ bracis-33639 128 9 node node ADJ bracis-33639 128 10 neighborhoods neighborhood NOUN bracis-33639 128 11 in in ADP bracis-33639 128 12 each each DET bracis-33639 128 13 snapshot snapshot NOUN bracis-33639 128 14 through through ADP bracis-33639 128 15 self self NOUN bracis-33639 128 16 - - PUNCT bracis-33639 128 17 attentional attentional ADJ bracis-33639 128 18 aggregation aggregation NOUN bracis-33639 128 19 . . PUNCT bracis-33639 129 1 the the DET bracis-33639 129 2 temporal temporal ADJ bracis-33639 129 3 attention attention NOUN bracis-33639 129 4 layer layer NOUN bracis-33639 129 5 captures capture VERB bracis-33639 129 6 the the DET bracis-33639 129 7 temporal temporal ADJ bracis-33639 129 8 variations variation NOUN bracis-33639 129 9 in in ADP bracis-33639 129 10 graph graph NOUN bracis-33639 129 11 structure structure NOUN bracis-33639 129 12 over over ADP bracis-33639 129 13 multiple multiple ADJ bracis-33639 129 14 time time NOUN bracis-33639 129 15 steps step NOUN bracis-33639 129 16 . . PUNCT bracis-33639 130 1 dysat dysat NOUN bracis-33639 130 2 leverages leverage NOUN bracis-33639 130 3 the the DET bracis-33639 130 4 self self NOUN bracis-33639 130 5 - - PUNCT bracis-33639 130 6 attention attention NOUN bracis-33639 130 7 mechanism mechanism NOUN bracis-33639 130 8 to to PART bracis-33639 130 9 concentrate concentrate VERB bracis-33639 130 10 on on ADP bracis-33639 130 11 the the DET bracis-33639 130 12 most most ADV bracis-33639 130 13 appropriate appropriate ADJ bracis-33639 130 14 parts part NOUN bracis-33639 130 15 of of ADP bracis-33639 130 16 the the DET bracis-33639 130 17 graph graph NOUN bracis-33639 130 18 when when SCONJ bracis-33639 130 19 updating update VERB bracis-33639 130 20 node node ADJ bracis-33639 130 21 representations representation NOUN bracis-33639 130 22 . . PUNCT bracis-33639 131 1 this this PRON bracis-33639 131 2 permits permit VERB bracis-33639 131 3 the the DET bracis-33639 131 4 model model NOUN bracis-33639 131 5 to to PART bracis-33639 131 6 effectively effectively ADV bracis-33639 131 7 consider consider VERB bracis-33639 131 8 the the DET bracis-33639 131 9 importance importance NOUN bracis-33639 131 10 of of ADP bracis-33639 131 11 different different ADJ bracis-33639 131 12 nodes node NOUN bracis-33639 131 13 and and CCONJ bracis-33639 131 14 edges edge NOUN bracis-33639 131 15 , , PUNCT bracis-33639 131 16 improving improve VERB bracis-33639 131 17 its its PRON bracis-33639 131 18 ability ability NOUN bracis-33639 131 19 to to PART bracis-33639 131 20 learn learn VERB bracis-33639 131 21 complex complex ADJ bracis-33639 131 22 patterns pattern NOUN bracis-33639 131 23 . . PUNCT bracis-33639 132 1 dysat dysat VERB bracis-33639 132 2 jointly jointly ADV bracis-33639 132 3 learns learn VERB bracis-33639 132 4 both both CCONJ bracis-33639 132 5 spatial spatial ADJ bracis-33639 132 6 ( ( PUNCT bracis-33639 132 7 structural structural ADJ bracis-33639 132 8 ) ) PUNCT bracis-33639 132 9 and and CCONJ bracis-33639 132 10 temporal temporal ADJ bracis-33639 132 11 patterns pattern NOUN bracis-33639 132 12 . . PUNCT bracis-33639 133 1 by by ADP bracis-33639 133 2 integrating integrate VERB bracis-33639 133 3 these these DET bracis-33639 133 4 two two NUM bracis-33639 133 5 types type NOUN bracis-33639 133 6 of of ADP bracis-33639 133 7 information information NOUN bracis-33639 133 8 , , PUNCT bracis-33639 133 9 it it PRON bracis-33639 133 10 constructs construct VERB bracis-33639 133 11 more more ADV bracis-33639 133 12 accurate accurate ADJ bracis-33639 133 13 and and CCONJ bracis-33639 133 14 robust robust ADJ bracis-33639 133 15 node node ADJ bracis-33639 133 16 embeddings embedding NOUN bracis-33639 133 17 . . PUNCT bracis-33639 134 1 the the DET bracis-33639 134 2 architecture architecture NOUN bracis-33639 134 3 of of ADP bracis-33639 134 4 the the DET bracis-33639 134 5 model model NOUN bracis-33639 134 6 is be AUX bracis-33639 134 7 as as SCONJ bracis-33639 134 8 follows follow VERB bracis-33639 134 9 : : PUNCT bracis-33639 134 10 input input NOUN bracis-33639 134 11 layer layer NOUN bracis-33639 134 12 : : PUNCT bracis-33639 134 13 the the DET bracis-33639 134 14 input input NOUN bracis-33639 134 15 to to ADP bracis-33639 134 16 the the DET bracis-33639 134 17 dysat dysat NOUN bracis-33639 134 18 model model NOUN bracis-33639 134 19 consists consist VERB bracis-33639 134 20 of of ADP bracis-33639 134 21 snapshots snapshot NOUN bracis-33639 134 22 of of ADP bracis-33639 134 23 the the DET bracis-33639 134 24 dynamic dynamic ADJ bracis-33639 134 25 graph graph NOUN bracis-33639 134 26 over over ADP bracis-33639 134 27 all all DET bracis-33639 134 28 time time NOUN bracis-33639 134 29 steps step NOUN bracis-33639 134 30 . . PUNCT bracis-33639 135 1 each each DET bracis-33639 135 2 snapshot snapshot NOUN bracis-33639 135 3 describes describe VERB bracis-33639 135 4 the the DET bracis-33639 135 5 graph graph NOUN bracis-33639 135 6 at at ADP bracis-33639 135 7 a a DET bracis-33639 135 8 particular particular ADJ bracis-33639 135 9 time time NOUN bracis-33639 135 10 step step NOUN bracis-33639 135 11 , , PUNCT bracis-33639 135 12 with with ADP bracis-33639 135 13 nodes node NOUN bracis-33639 135 14 and and CCONJ bracis-33639 135 15 edges edge NOUN bracis-33639 135 16 having have VERB bracis-33639 135 17 associated associate VERB bracis-33639 135 18 features feature NOUN bracis-33639 135 19 . . PUNCT bracis-33639 136 1 spatial spatial ADJ bracis-33639 136 2 self self NOUN bracis-33639 136 3 - - PUNCT bracis-33639 136 4 attention attention NOUN bracis-33639 136 5 : : PUNCT bracis-33639 136 6 each each DET bracis-33639 136 7 snapshot snapshot NOUN bracis-33639 136 8 undergoes undergo VERB bracis-33639 136 9 a a DET bracis-33639 136 10 spatial spatial ADJ bracis-33639 136 11 self self NOUN bracis-33639 136 12 - - PUNCT bracis-33639 136 13 attention attention NOUN bracis-33639 136 14 layer layer NOUN bracis-33639 136 15 to to PART bracis-33639 136 16 capture capture VERB bracis-33639 136 17 structural structural ADJ bracis-33639 136 18 associations association NOUN bracis-33639 136 19 , , PUNCT bracis-33639 136 20 estimating estimate VERB bracis-33639 136 21 attention attention NOUN bracis-33639 136 22 scores score NOUN bracis-33639 136 23 between between ADP bracis-33639 136 24 nodes node NOUN bracis-33639 136 25 to to PART bracis-33639 136 26 aggregate aggregate VERB bracis-33639 136 27 information information NOUN bracis-33639 136 28 from from ADP bracis-33639 136 29 relevant relevant ADJ bracis-33639 136 30 neighbors neighbor NOUN bracis-33639 136 31 . . PUNCT bracis-33639 137 1 temporal temporal ADJ bracis-33639 137 2 self self NOUN bracis-33639 137 3 - - PUNCT bracis-33639 137 4 attention attention NOUN bracis-33639 137 5 : : PUNCT bracis-33639 137 6 spatial spatial ADJ bracis-33639 137 7 embeddings embedding NOUN bracis-33639 137 8 are be AUX bracis-33639 137 9 then then ADV bracis-33639 137 10 processed process VERB bracis-33639 137 11 through through ADP bracis-33639 137 12 a a DET bracis-33639 137 13 temporal temporal ADJ bracis-33639 137 14 self self NOUN bracis-33639 137 15 - - PUNCT bracis-33639 137 16 attention attention NOUN bracis-33639 137 17 layer layer NOUN bracis-33639 137 18 to to PART bracis-33639 137 19 learn learn VERB bracis-33639 137 20 temporal temporal ADJ bracis-33639 137 21 dependencies dependency NOUN bracis-33639 137 22 and and CCONJ bracis-33639 137 23 how how SCONJ bracis-33639 137 24 node node ADJ bracis-33639 137 25 representations representation NOUN bracis-33639 137 26 evolve evolve VERB bracis-33639 137 27 over over ADP bracis-33639 137 28 time time NOUN bracis-33639 137 29 . . PUNCT bracis-33639 138 1 output output NOUN bracis-33639 138 2 layer layer NOUN bracis-33639 138 3 : : PUNCT bracis-33639 138 4 final final ADJ bracis-33639 138 5 node node ADJ bracis-33639 138 6 embeddings embedding NOUN bracis-33639 138 7 are be AUX bracis-33639 138 8 obtained obtain VERB bracis-33639 138 9 by by ADP bracis-33639 138 10 integrating integrate VERB bracis-33639 138 11 spatial spatial ADJ bracis-33639 138 12 and and CCONJ bracis-33639 138 13 temporal temporal ADJ bracis-33639 138 14 embeddings embedding NOUN bracis-33639 138 15 , , PUNCT bracis-33639 138 16 used use VERB bracis-33639 138 17 for for ADP bracis-33639 138 18 tasks task NOUN bracis-33639 138 19 like like ADP bracis-33639 138 20 node node NOUN bracis-33639 138 21 classification classification NOUN bracis-33639 138 22 , , PUNCT bracis-33639 138 23 link link NOUN bracis-33639 138 24 prediction prediction NOUN bracis-33639 138 25 , , PUNCT bracis-33639 138 26 and and CCONJ bracis-33639 138 27 anomaly anomaly NOUN bracis-33639 138 28 detection detection NOUN bracis-33639 138 29 . . PUNCT bracis-33639 139 1 in in ADP bracis-33639 139 2 this this DET bracis-33639 139 3 study study NOUN bracis-33639 139 4 , , PUNCT bracis-33639 139 5 we we PRON bracis-33639 139 6 used use VERB bracis-33639 139 7 them they PRON bracis-33639 139 8 for for ADP bracis-33639 139 9 node node ADJ bracis-33639 139 10 classification classification NOUN bracis-33639 139 11 . . PUNCT bracis-33639 140 1 4.2 4.2 NUM bracis-33639 140 2 evolvegcn evolvegcn ADV bracis-33639 140 3 the the DET bracis-33639 140 4 main main ADJ bracis-33639 140 5 goal goal NOUN bracis-33639 140 6 of of ADP bracis-33639 140 7 evolving evolve VERB bracis-33639 140 8 graph graph NOUN bracis-33639 140 9 convolutional convolutional ADJ bracis-33639 140 10 networks network NOUN bracis-33639 140 11 ( ( PUNCT bracis-33639 140 12 evolvegcn evolvegcn PROPN bracis-33639 140 13 ) ) PUNCT bracis-33639 141 1 [ [ X bracis-33639 141 2 22 22 NUM bracis-33639 141 3 ] ] PUNCT bracis-33639 141 4 is be AUX bracis-33639 141 5 to to PART bracis-33639 141 6 adapt adapt VERB bracis-33639 141 7 gcns gcns PROPN bracis-33639 141 8 , , PUNCT bracis-33639 141 9 originally originally ADV bracis-33639 141 10 designed design VERB bracis-33639 141 11 for for ADP bracis-33639 141 12 static static ADJ bracis-33639 141 13 graphs graph NOUN bracis-33639 141 14 , , PUNCT bracis-33639 141 15 to to PART bracis-33639 141 16 effectively effectively ADV bracis-33639 141 17 handle handle VERB bracis-33639 141 18 graphs graph NOUN bracis-33639 141 19 that that PRON bracis-33639 141 20 evolve evolve VERB bracis-33639 141 21 continuously continuously ADV bracis-33639 141 22 . . PUNCT bracis-33639 142 1 this this DET bracis-33639 142 2 adaptation adaptation NOUN bracis-33639 142 3 is be AUX bracis-33639 142 4 crucial crucial ADJ bracis-33639 142 5 for for ADP bracis-33639 142 6 real real ADJ bracis-33639 142 7 - - PUNCT bracis-33639 142 8 world world NOUN bracis-33639 142 9 applications application NOUN bracis-33639 142 10 such such ADJ bracis-33639 142 11 as as ADP bracis-33639 142 12 social social ADJ bracis-33639 142 13 networks network NOUN bracis-33639 142 14 , , PUNCT bracis-33639 142 15 recommendation recommendation NOUN bracis-33639 142 16 systems system NOUN bracis-33639 142 17 , , PUNCT bracis-33639 142 18 and and CCONJ bracis-33639 142 19 communication communication NOUN bracis-33639 142 20 networks network NOUN bracis-33639 142 21 , , PUNCT bracis-33639 142 22 where where SCONJ bracis-33639 142 23 relationships relationship NOUN bracis-33639 142 24 between between ADP bracis-33639 142 25 entities entity NOUN bracis-33639 142 26 evolve evolve VERB bracis-33639 142 27 over over ADP bracis-33639 142 28 time time NOUN bracis-33639 142 29 . . PUNCT bracis-33639 143 1 evolvegcn evolvegcn PROPN bracis-33639 143 2 achieves achieve VERB bracis-33639 143 3 this this PRON bracis-33639 143 4 by by ADP bracis-33639 143 5 introducing introduce VERB bracis-33639 143 6 an an DET bracis-33639 143 7 evolutionary evolutionary ADJ bracis-33639 143 8 approach approach NOUN bracis-33639 143 9 to to PART bracis-33639 143 10 update update VERB bracis-33639 143 11 gcn gcn NOUN bracis-33639 143 12 parameters parameter NOUN bracis-33639 143 13 as as SCONJ bracis-33639 143 14 the the DET bracis-33639 143 15 graph graph NOUN bracis-33639 143 16 evolves evolve VERB bracis-33639 143 17 . . PUNCT bracis-33639 144 1 it it PRON bracis-33639 144 2 treats treat VERB bracis-33639 144 3 these these DET bracis-33639 144 4 parameters parameter NOUN bracis-33639 144 5 as as ADP bracis-33639 144 6 temporal temporal ADJ bracis-33639 144 7 sequences sequence NOUN bracis-33639 144 8 and and CCONJ bracis-33639 144 9 utilizes utilize VERB bracis-33639 144 10 time time NOUN bracis-33639 144 11 - - PUNCT bracis-33639 144 12 series series NOUN bracis-33639 144 13 models model NOUN bracis-33639 144 14 to to PART bracis-33639 144 15 predict predict VERB bracis-33639 144 16 and and CCONJ bracis-33639 144 17 adjust adjust VERB bracis-33639 144 18 them they PRON bracis-33639 144 19 over over ADP bracis-33639 144 20 time time NOUN bracis-33639 144 21 . . PUNCT bracis-33639 145 1 the the DET bracis-33639 145 2 method method NOUN bracis-33639 145 3 focuses focus VERB bracis-33639 145 4 on on ADP bracis-33639 145 5 two two NUM bracis-33639 145 6 primary primary ADJ bracis-33639 145 7 models model NOUN bracis-33639 145 8 for for ADP bracis-33639 145 9 parameter parameter NOUN bracis-33639 145 10 evolution evolution NOUN bracis-33639 145 11 : : PUNCT bracis-33639 145 12 hidden hide VERB bracis-33639 145 13 model model NOUN bracis-33639 145 14 : : PUNCT bracis-33639 145 15 this this DET bracis-33639 145 16 model model NOUN bracis-33639 145 17 employs employ VERB bracis-33639 145 18 lstm lstm ADJ bracis-33639 145 19 networks network NOUN bracis-33639 145 20 to to PART bracis-33639 145 21 capture capture VERB bracis-33639 145 22 temporal temporal ADJ bracis-33639 145 23 dependencies dependency NOUN bracis-33639 145 24 and and CCONJ bracis-33639 145 25 evolve evolve VERB bracis-33639 145 26 gcn gcn NOUN bracis-33639 145 27 parameters parameter NOUN bracis-33639 145 28 based base VERB bracis-33639 145 29 on on ADP bracis-33639 145 30 historical historical ADJ bracis-33639 145 31 states state NOUN bracis-33639 145 32 . . PUNCT bracis-33639 146 1 output output NOUN bracis-33639 146 2 model model PROPN bracis-33639 146 3 : : PUNCT bracis-33639 146 4 in in ADP bracis-33639 146 5 contrast contrast NOUN bracis-33639 146 6 , , PUNCT bracis-33639 146 7 this this DET bracis-33639 146 8 model model NOUN bracis-33639 146 9 uses use VERB bracis-33639 146 10 direct direct ADJ bracis-33639 146 11 rnn rnn NOUN bracis-33639 146 12 architectures architecture NOUN bracis-33639 146 13 to to PART bracis-33639 146 14 update update VERB bracis-33639 146 15 parameters parameter NOUN bracis-33639 146 16 based base VERB bracis-33639 146 17 on on ADP bracis-33639 146 18 output output NOUN bracis-33639 146 19 states state NOUN bracis-33639 146 20 from from ADP bracis-33639 146 21 the the DET bracis-33639 146 22 previous previous ADJ bracis-33639 146 23 layers layer NOUN bracis-33639 146 24 . . PUNCT bracis-33639 147 1 in in ADP bracis-33639 147 2 this this DET bracis-33639 147 3 work work NOUN bracis-33639 147 4 , , PUNCT bracis-33639 147 5 we we PRON bracis-33639 147 6 make make VERB bracis-33639 147 7 use use NOUN bracis-33639 147 8 of of ADP bracis-33639 147 9 the the DET bracis-33639 147 10 hidden hide VERB bracis-33639 147 11 version version NOUN bracis-33639 147 12 . . PUNCT bracis-33639 148 1 the the DET bracis-33639 148 2 training training NOUN bracis-33639 148 3 of of ADP bracis-33639 148 4 evolvegcn evolvegcn PROPN bracis-33639 148 5 involves involve VERB bracis-33639 148 6 optimizing optimize VERB bracis-33639 148 7 not not PART bracis-33639 148 8 only only ADV bracis-33639 148 9 the the DET bracis-33639 148 10 base base NOUN bracis-33639 148 11 gcn gcn NOUN bracis-33639 148 12 ’s ’s PART bracis-33639 148 13 parameters parameter NOUN bracis-33639 148 14 but but CCONJ bracis-33639 148 15 also also ADV bracis-33639 148 16 the the DET bracis-33639 148 17 evolution evolution PROPN bracis-33639 148 18 model model PROPN bracis-33639 148 19 ’s ’s PART bracis-33639 148 20 ( ( PUNCT bracis-33639 148 21 lstm lstm NOUN bracis-33639 148 22 or or CCONJ bracis-33639 148 23 rnn rnn NOUN bracis-33639 148 24 ) ) PUNCT bracis-33639 148 25 parameters parameter NOUN bracis-33639 148 26 . . PUNCT bracis-33639 149 1 at at ADP bracis-33639 149 2 each each DET bracis-33639 149 3 new new ADJ bracis-33639 149 4 time time NOUN bracis-33639 149 5 step step NOUN bracis-33639 149 6 , , PUNCT bracis-33639 149 7 the the DET bracis-33639 149 8 gcn gcn NOUN bracis-33639 149 9 ’s ’s PART bracis-33639 149 10 weights weight NOUN bracis-33639 149 11 are be AUX bracis-33639 149 12 updated update VERB bracis-33639 149 13 using use VERB bracis-33639 149 14 the the DET bracis-33639 149 15 selected select VERB bracis-33639 149 16 evolution evolution NOUN bracis-33639 149 17 model model NOUN bracis-33639 149 18 , , PUNCT bracis-33639 149 19 ensuring ensure VERB bracis-33639 149 20 that that SCONJ bracis-33639 149 21 the the DET bracis-33639 149 22 network network NOUN bracis-33639 149 23 adapts adapt VERB bracis-33639 149 24 to to ADP bracis-33639 149 25 the the DET bracis-33639 149 26 graph graph NOUN bracis-33639 149 27 ’s ’s PART bracis-33639 149 28 changing change VERB bracis-33639 149 29 dynamics dynamic NOUN bracis-33639 149 30 . . PUNCT bracis-33639 150 1 4.3 4.3 NUM bracis-33639 150 2 baseline baseline NOUN bracis-33639 150 3 models model NOUN bracis-33639 150 4 we we PRON bracis-33639 150 5 compare compare VERB bracis-33639 150 6 the the DET bracis-33639 150 7 gnn gnn NOUN bracis-33639 150 8 models model NOUN bracis-33639 150 9 against against ADP bracis-33639 150 10 three three NUM bracis-33639 150 11 traditional traditional ADJ bracis-33639 150 12 non non ADJ bracis-33639 150 13 - - ADJ bracis-33639 150 14 deep deep ADJ bracis-33639 150 15 learning learning NOUN bracis-33639 150 16 models model NOUN bracis-33639 150 17 : : PUNCT bracis-33639 150 18 support support NOUN bracis-33639 150 19 vector vector NOUN bracis-33639 150 20 machine machine NOUN bracis-33639 150 21 ( ( PUNCT bracis-33639 150 22 svm svm PROPN bracis-33639 150 23 ) ) PUNCT bracis-33639 150 24 , , PUNCT bracis-33639 150 25 random random ADJ bracis-33639 150 26 forest forest NOUN bracis-33639 150 27 and and CCONJ bracis-33639 150 28 logistic logistic ADJ bracis-33639 150 29 regression regression NOUN bracis-33639 150 30 . . PUNCT bracis-33639 151 1 this this DET bracis-33639 151 2 analysis analysis NOUN bracis-33639 151 3 aims aim VERB bracis-33639 151 4 to to PART bracis-33639 151 5 highlight highlight VERB bracis-33639 151 6 the the DET bracis-33639 151 7 predictive predictive ADJ bracis-33639 151 8 capabilities capability NOUN bracis-33639 151 9 of of ADP bracis-33639 151 10 gnns gnns NOUN bracis-33639 151 11 in in ADP bracis-33639 151 12 contrast contrast NOUN bracis-33639 151 13 to to ADP bracis-33639 151 14 these these DET bracis-33639 151 15 established establish VERB bracis-33639 151 16 machine machine NOUN bracis-33639 151 17 learning learning NOUN bracis-33639 151 18 approaches approach NOUN bracis-33639 151 19 . . PUNCT bracis-33639 152 1 by by ADP bracis-33639 152 2 evaluating evaluate VERB bracis-33639 152 3 the the DET bracis-33639 152 4 performance performance NOUN bracis-33639 152 5 of of ADP bracis-33639 152 6 these these DET bracis-33639 152 7 traditional traditional ADJ bracis-33639 152 8 models model NOUN bracis-33639 152 9 , , PUNCT bracis-33639 152 10 we we PRON bracis-33639 152 11 provide provide VERB bracis-33639 152 12 a a DET bracis-33639 152 13 benchmark benchmark NOUN bracis-33639 152 14 to to PART bracis-33639 152 15 better well ADV bracis-33639 152 16 understand understand VERB bracis-33639 152 17 the the DET bracis-33639 152 18 advantages advantage NOUN bracis-33639 152 19 and and CCONJ bracis-33639 152 20 limitations limitation NOUN bracis-33639 152 21 of of ADP bracis-33639 152 22 gnns gnns NOUN bracis-33639 152 23 in in ADP bracis-33639 152 24 the the DET bracis-33639 152 25 context context NOUN bracis-33639 152 26 of of ADP bracis-33639 152 27 crime crime NOUN bracis-33639 152 28 prediction prediction NOUN bracis-33639 152 29 . . PUNCT bracis-33639 153 1 5 5 NUM bracis-33639 153 2 experimental experimental ADJ bracis-33639 153 3 setup setup NOUN bracis-33639 153 4 we we PRON bracis-33639 153 5 conducted conduct VERB bracis-33639 153 6 experiments experiment NOUN bracis-33639 153 7 on on ADP bracis-33639 153 8 a a DET bracis-33639 153 9 real real ADJ bracis-33639 153 10 - - PUNCT bracis-33639 153 11 world world NOUN bracis-33639 153 12 crime crime NOUN bracis-33639 153 13 records record NOUN bracis-33639 153 14 dataset dataset VERB bracis-33639 153 15 collected collect VERB bracis-33639 153 16 in in ADP bracis-33639 153 17 sao sao PROPN bracis-33639 153 18 paulo paulo PROPN bracis-33639 153 19 from from ADP bracis-33639 153 20 jan jan PROPN bracis-33639 153 21 . . PROPN bracis-33639 153 22 1st 1st NOUN bracis-33639 153 23 , , PUNCT bracis-33639 153 24 2021 2021 NUM bracis-33639 153 25 , , PUNCT bracis-33639 153 26 to to ADP bracis-33639 153 27 dec dec PROPN bracis-33639 153 28 . . PROPN bracis-33639 153 29 31st 31st NOUN bracis-33639 153 30 , , PUNCT bracis-33639 153 31 2022 2022 NUM bracis-33639 153 32 . . PUNCT bracis-33639 154 1 we we PRON bracis-33639 154 2 integrated integrate VERB bracis-33639 154 3 the the DET bracis-33639 154 4 crime crime NOUN bracis-33639 154 5 data datum NOUN bracis-33639 154 6 with with ADP bracis-33639 154 7 street street NOUN bracis-33639 154 8 map map NOUN bracis-33639 154 9 data datum NOUN bracis-33639 154 10 and and CCONJ bracis-33639 154 11 transformed transform VERB bracis-33639 154 12 them they PRON bracis-33639 154 13 into into ADP bracis-33639 154 14 dynamic dynamic ADJ bracis-33639 154 15 graphs graph NOUN bracis-33639 154 16 ; ; PUNCT bracis-33639 154 17 the the DET bracis-33639 154 18 transformation transformation NOUN bracis-33639 154 19 details detail NOUN bracis-33639 154 20 are be AUX bracis-33639 154 21 already already ADV bracis-33639 154 22 discussed discuss VERB bracis-33639 154 23 in in ADP bracis-33639 154 24 sect sect NOUN bracis-33639 154 25 . . PUNCT bracis-33639 154 26     SPACE bracis-33639 155 1 3 3 X bracis-33639 155 2 . . X bracis-33639 155 3 dynamic dynamic ADJ bracis-33639 155 4 graphs graph NOUN bracis-33639 155 5 are be AUX bracis-33639 155 6 a a DET bracis-33639 155 7 series series NOUN bracis-33639 155 8 of of ADP bracis-33639 155 9 static static ADJ bracis-33639 155 10 graph graph NOUN bracis-33639 155 11 snapshots snapshot NOUN bracis-33639 155 12 , , PUNCT bracis-33639 155 13 \(g \(g NOUN bracis-33639 155 14 = = SYM bracis-33639 155 15 \{g^1 \{g^1 PROPN bracis-33639 155 16 , , PUNCT bracis-33639 155 17 ... ... PUNCT bracis-33639 155 18 , , PUNCT bracis-33639 155 19 g^t\}\ g^t\}\ NOUN bracis-33639 155 20 ) ) PUNCT bracis-33639 155 21 where where SCONJ bracis-33639 155 22 t t PROPN bracis-33639 155 23 represents represent VERB bracis-33639 155 24 the the DET bracis-33639 155 25 number number NOUN bracis-33639 155 26 of of ADP bracis-33639 155 27 time time NOUN bracis-33639 155 28 steps step NOUN bracis-33639 155 29 . . PUNCT bracis-33639 156 1 in in ADP bracis-33639 156 2 this this DET bracis-33639 156 3 work work NOUN bracis-33639 156 4 , , PUNCT bracis-33639 156 5 each each DET bracis-33639 156 6 snapshot snapshot NOUN bracis-33639 156 7 represents represent VERB bracis-33639 156 8 the the DET bracis-33639 156 9 crime crime NOUN bracis-33639 156 10 data datum NOUN bracis-33639 156 11 for for ADP bracis-33639 156 12 a a DET bracis-33639 156 13 month month NOUN bracis-33639 156 14 , , PUNCT bracis-33639 156 15 therefore therefore ADV bracis-33639 156 16 , , PUNCT bracis-33639 156 17 in in ADP bracis-33639 156 18 total total NOUN bracis-33639 156 19 , , PUNCT bracis-33639 156 20 we we PRON bracis-33639 156 21 have have VERB bracis-33639 156 22 24 24 NUM bracis-33639 156 23 months month NOUN bracis-33639 156 24 of of ADP bracis-33639 156 25 graph graph NOUN bracis-33639 156 26 snapshots snapshot NOUN bracis-33639 156 27 . . PUNCT bracis-33639 157 1 we we PRON bracis-33639 157 2 generated generate VERB bracis-33639 157 3 22 22 NUM bracis-33639 157 4 instances instance NOUN bracis-33639 157 5 from from ADP bracis-33639 157 6 24 24 NUM bracis-33639 157 7 snapshots snapshot NOUN bracis-33639 157 8 using use VERB bracis-33639 157 9 a a DET bracis-33639 157 10 sliding slide VERB bracis-33639 157 11 window window NOUN bracis-33639 157 12 approach approach NOUN bracis-33639 157 13 , , PUNCT bracis-33639 157 14 with with ADP bracis-33639 157 15 each each DET bracis-33639 157 16 instance instance NOUN bracis-33639 157 17 comprising comprise VERB bracis-33639 157 18 three three NUM bracis-33639 157 19 consecutive consecutive ADJ bracis-33639 157 20 months month NOUN bracis-33639 157 21 ( ( PUNCT bracis-33639 157 22 the the DET bracis-33639 157 23 first first ADJ bracis-33639 157 24 two two NUM bracis-33639 157 25 for for ADP bracis-33639 157 26 training training NOUN bracis-33639 157 27 and and CCONJ bracis-33639 157 28 the the DET bracis-33639 157 29 last last ADJ bracis-33639 157 30 for for ADP bracis-33639 157 31 evaluation evaluation NOUN bracis-33639 157 32 ) ) PUNCT bracis-33639 157 33 . . PUNCT bracis-33639 158 1 these these PRON bracis-33639 158 2 were be AUX bracis-33639 158 3 split split VERB bracis-33639 158 4 80 80 NUM bracis-33639 158 5 - - SYM bracis-33639 158 6 20 20 NUM bracis-33639 158 7 ratio ratio NOUN bracis-33639 158 8 into into ADP bracis-33639 158 9 17 17 NUM bracis-33639 158 10 for for ADP bracis-33639 158 11 train train NOUN bracis-33639 158 12 / / SYM bracis-33639 158 13 validation validation NOUN bracis-33639 158 14 and and CCONJ bracis-33639 158 15 5 5 NUM bracis-33639 158 16 for for ADP bracis-33639 158 17 testing testing NOUN bracis-33639 158 18 , , PUNCT bracis-33639 158 19 with with SCONJ bracis-33639 158 20 the the DET bracis-33639 158 21 train train NOUN bracis-33639 158 22 / / SYM bracis-33639 158 23 validation validation NOUN bracis-33639 158 24 set set VERB bracis-33639 158 25 further far ADV bracis-33639 158 26 divided divide VERB bracis-33639 158 27 into into ADP bracis-33639 158 28 12 12 NUM bracis-33639 158 29 for for ADP bracis-33639 158 30 training training NOUN bracis-33639 158 31 and and CCONJ bracis-33639 158 32 5 5 NUM bracis-33639 158 33 for for ADP bracis-33639 158 34 validation validation NOUN bracis-33639 158 35 . . PUNCT bracis-33639 159 1 this this DET bracis-33639 159 2 method method NOUN bracis-33639 159 3 captures capture VERB bracis-33639 159 4 temporal temporal ADJ bracis-33639 159 5 dependencies dependency NOUN bracis-33639 159 6 while while SCONJ bracis-33639 159 7 ensuring ensure VERB bracis-33639 159 8 unbiased unbiased ADJ bracis-33639 159 9 evaluation evaluation NOUN bracis-33639 159 10 . . PUNCT bracis-33639 160 1 each each DET bracis-33639 160 2 graph graph NOUN bracis-33639 160 3 has have VERB bracis-33639 160 4 4021 4021 NUM bracis-33639 160 5 nodes node NOUN bracis-33639 160 6 and and CCONJ bracis-33639 160 7 19899 19899 NUM bracis-33639 160 8 edges edge NOUN bracis-33639 160 9 . . PUNCT bracis-33639 161 1 the the DET bracis-33639 161 2 graph graph NOUN bracis-33639 161 3 statistics statistic NOUN bracis-33639 161 4 are be AUX bracis-33639 161 5 described describe VERB bracis-33639 161 6 in in ADP bracis-33639 161 7 table table NOUN bracis-33639 161 8     SPACE bracis-33639 161 9 1 1 NUM bracis-33639 161 10 . . PUNCT bracis-33639 161 11 table table NOUN bracis-33639 161 12 1 1 NUM bracis-33639 161 13 . . PUNCT bracis-33639 161 14 crime crime NOUN bracis-33639 161 15 dataset dataset VERB bracis-33639 161 16 summary.full summary.full NOUN bracis-33639 161 17 size size NOUN bracis-33639 161 18 table table NOUN bracis-33639 161 19 5.1 5.1 NUM bracis-33639 161 20 models model NOUN bracis-33639 161 21 hyperparameters hyperparameter VERB bracis-33639 161 22 we we PRON bracis-33639 161 23 conducted conduct VERB bracis-33639 161 24 several several ADJ bracis-33639 161 25 experiments experiment NOUN bracis-33639 161 26 varying vary VERB bracis-33639 161 27 the the DET bracis-33639 161 28 hyperparameters hyperparameter NOUN bracis-33639 161 29 and and CCONJ bracis-33639 161 30 selected select VERB bracis-33639 161 31 the the DET bracis-33639 161 32 most most ADV bracis-33639 161 33 advantageous advantageous ADJ bracis-33639 161 34 trade trade NOUN bracis-33639 161 35 - - PUNCT bracis-33639 161 36 off off NOUN bracis-33639 161 37 among among ADP bracis-33639 161 38 the the DET bracis-33639 161 39 metrics metric NOUN bracis-33639 161 40 used use VERB bracis-33639 161 41 . . PUNCT bracis-33639 162 1 ultimately ultimately ADV bracis-33639 162 2 , , PUNCT bracis-33639 162 3 we we PRON bracis-33639 162 4 chose choose VERB bracis-33639 162 5 the the DET bracis-33639 162 6 hyperparameters hyperparameter NOUN bracis-33639 162 7 based base VERB bracis-33639 162 8 on on ADP bracis-33639 162 9 the the DET bracis-33639 162 10 stability stability NOUN bracis-33639 162 11 of of ADP bracis-33639 162 12 the the DET bracis-33639 162 13 performance performance NOUN bracis-33639 162 14 in in ADP bracis-33639 162 15 the the DET bracis-33639 162 16 training training NOUN bracis-33639 162 17 and and CCONJ bracis-33639 162 18 validation validation NOUN bracis-33639 162 19 sets set NOUN bracis-33639 162 20 , , PUNCT bracis-33639 162 21 which which PRON bracis-33639 162 22 led lead VERB bracis-33639 162 23 to to ADP bracis-33639 162 24 a a DET bracis-33639 162 25 significant significant ADJ bracis-33639 162 26 improvement improvement NOUN bracis-33639 162 27 . . PUNCT bracis-33639 163 1 dysat dysat PROPN bracis-33639 163 2 . . PUNCT bracis-33639 164 1 in in ADP bracis-33639 164 2 our our PRON bracis-33639 164 3 dysat dysat NOUN bracis-33639 164 4 model model NOUN bracis-33639 164 5 experiments experiment NOUN bracis-33639 164 6 , , PUNCT bracis-33639 164 7 we we PRON bracis-33639 164 8 carefully carefully ADV bracis-33639 164 9 selected select VERB bracis-33639 164 10 hyperparameters hyperparameter NOUN bracis-33639 164 11 to to PART bracis-33639 164 12 balance balance VERB bracis-33639 164 13 complexity complexity NOUN bracis-33639 164 14 , , PUNCT bracis-33639 164 15 efficiency efficiency NOUN bracis-33639 164 16 , , PUNCT bracis-33639 164 17 and and CCONJ bracis-33639 164 18 overfitting overfitte VERB bracis-33639 164 19 . . PUNCT bracis-33639 165 1 to to PART bracis-33639 165 2 address address VERB bracis-33639 165 3 the the DET bracis-33639 165 4 dataset dataset NOUN bracis-33639 165 5 ’s ’s PART bracis-33639 165 6 imbalance imbalance NOUN bracis-33639 165 7 , , PUNCT bracis-33639 165 8 we we PRON bracis-33639 165 9 used use VERB bracis-33639 165 10 classweights classweight NOUN bracis-33639 165 11 with with ADP bracis-33639 165 12 a a DET bracis-33639 165 13 1:4 1:4 NUM bracis-33639 165 14 ratio ratio NOUN bracis-33639 165 15 for for ADP bracis-33639 165 16 majority majority NOUN bracis-33639 165 17 and and CCONJ bracis-33639 165 18 minority minority NOUN bracis-33639 165 19 classes class NOUN bracis-33639 165 20 . . PUNCT bracis-33639 166 1 we we PRON bracis-33639 166 2 set set VERB bracis-33639 166 3 the the DET bracis-33639 166 4 learning learning NOUN bracis-33639 166 5 rate rate NOUN bracis-33639 166 6 at at ADP bracis-33639 166 7 0.01 0.01 NUM bracis-33639 166 8 and and CCONJ bracis-33639 166 9 applied apply VERB bracis-33639 166 10 weight weight NOUN bracis-33639 166 11 decay decay NOUN bracis-33639 166 12 of of ADP bracis-33639 166 13 0.0005 0.0005 NUM bracis-33639 166 14 to to PART bracis-33639 166 15 regularize regularize VERB bracis-33639 166 16 the the DET bracis-33639 166 17 model model NOUN bracis-33639 166 18 . . PUNCT bracis-33639 167 1 dropout dropout NOUN bracis-33639 167 2 rates rate NOUN bracis-33639 167 3 of of ADP bracis-33639 167 4 0.4 0.4 NUM bracis-33639 167 5 and and CCONJ bracis-33639 167 6 0.5 0.5 NUM bracis-33639 167 7 were be AUX bracis-33639 167 8 used use VERB bracis-33639 167 9 for for ADP bracis-33639 167 10 spatial spatial ADJ bracis-33639 167 11 and and CCONJ bracis-33639 167 12 temporal temporal ADJ bracis-33639 167 13 attention attention NOUN bracis-33639 167 14 to to PART bracis-33639 167 15 enhance enhance VERB bracis-33639 167 16 robustness robustness NOUN bracis-33639 167 17 . . PUNCT bracis-33639 168 1 the the DET bracis-33639 168 2 model model NOUN bracis-33639 168 3 ’s ’s PART bracis-33639 168 4 architecture architecture NOUN bracis-33639 168 5 included include VERB bracis-33639 168 6 16 16 NUM bracis-33639 168 7 attention attention NOUN bracis-33639 168 8 heads head NOUN bracis-33639 168 9 and and CCONJ bracis-33639 168 10 128 128 NUM bracis-33639 168 11 units unit NOUN bracis-33639 168 12 in in ADP bracis-33639 168 13 spatial spatial ADJ bracis-33639 168 14 and and CCONJ bracis-33639 168 15 temporal temporal ADJ bracis-33639 168 16 layers layer NOUN bracis-33639 168 17 , , PUNCT bracis-33639 168 18 enabling enable VERB bracis-33639 168 19 it it PRON bracis-33639 168 20 to to PART bracis-33639 168 21 capture capture VERB bracis-33639 168 22 complex complex ADJ bracis-33639 168 23 relationships relationship NOUN bracis-33639 168 24 while while SCONJ bracis-33639 168 25 maintaining maintain VERB bracis-33639 168 26 computational computational ADJ bracis-33639 168 27 efficiency efficiency NOUN bracis-33639 168 28 . . PUNCT bracis-33639 169 1 table table NOUN bracis-33639 169 2 2 2 NUM bracis-33639 169 3 . . PUNCT bracis-33639 169 4 models model NOUN bracis-33639 169 5 hyperparameters.full hyperparameters.full PRON bracis-33639 169 6 size size NOUN bracis-33639 169 7 table table NOUN bracis-33639 169 8 evolvegcn evolvegcn NOUN bracis-33639 169 9 . . PUNCT bracis-33639 170 1 three three NUM bracis-33639 170 2 primary primary ADJ bracis-33639 170 3 hyperparameters hyperparameter NOUN bracis-33639 170 4 in in ADP bracis-33639 170 5 evolvegcn evolvegcn PROPN bracis-33639 170 6 significantly significantly ADV bracis-33639 170 7 impact impact VERB bracis-33639 170 8 its its PRON bracis-33639 170 9 performance performance NOUN bracis-33639 170 10 . . PUNCT bracis-33639 171 1 the the DET bracis-33639 171 2 first first ADJ bracis-33639 171 3 , , PUNCT bracis-33639 171 4 classweights classweight NOUN bracis-33639 171 5 , , PUNCT bracis-33639 171 6 assigns assign VERB bracis-33639 171 7 weights weight NOUN bracis-33639 171 8 to to ADP bracis-33639 171 9 classes class NOUN bracis-33639 171 10 , , PUNCT bracis-33639 171 11 with with ADP bracis-33639 171 12 the the DET bracis-33639 171 13 second second ADJ bracis-33639 171 14 value value NOUN bracis-33639 171 15 weighting weight VERB bracis-33639 171 16 the the DET bracis-33639 171 17 target target NOUN bracis-33639 171 18 class class NOUN bracis-33639 171 19 ( ( PUNCT bracis-33639 171 20 class class NOUN bracis-33639 171 21 1 1 NUM bracis-33639 171 22 ) ) PUNCT bracis-33639 171 23 and and CCONJ bracis-33639 171 24 the the DET bracis-33639 171 25 first first ADJ bracis-33639 171 26 value value NOUN bracis-33639 171 27 weighting weight VERB bracis-33639 171 28 class class NOUN bracis-33639 171 29 0 0 NUM bracis-33639 171 30 . . PUNCT bracis-33639 172 1 the the DET bracis-33639 172 2 second second ADJ bracis-33639 172 3 , , PUNCT bracis-33639 172 4 numberofsteps numberofstep NOUN bracis-33639 172 5 , , PUNCT bracis-33639 172 6 defines define VERB bracis-33639 172 7 the the DET bracis-33639 172 8 number number NOUN bracis-33639 172 9 of of ADP bracis-33639 172 10 temporal temporal ADJ bracis-33639 172 11 layers layer NOUN bracis-33639 172 12 ( ( PUNCT bracis-33639 172 13 e.g. e.g. ADV bracis-33639 172 14 , , PUNCT bracis-33639 172 15 months month NOUN bracis-33639 172 16 ) ) PUNCT bracis-33639 172 17 the the DET bracis-33639 172 18 model model NOUN bracis-33639 172 19 considers consider VERB bracis-33639 172 20 for for ADP bracis-33639 172 21 predictions prediction NOUN bracis-33639 172 22 . . PUNCT bracis-33639 173 1 the the DET bracis-33639 173 2 third third ADJ bracis-33639 173 3 , , PUNCT bracis-33639 173 4 featurespernodes featurespernode NOUN bracis-33639 173 5 , , PUNCT bracis-33639 173 6 determines determine VERB bracis-33639 173 7 the the DET bracis-33639 173 8 input input NOUN bracis-33639 173 9 and and CCONJ bracis-33639 173 10 output output NOUN bracis-33639 173 11 dimensions dimension NOUN bracis-33639 173 12 between between ADP bracis-33639 173 13 network network NOUN bracis-33639 173 14 layers layer NOUN bracis-33639 173 15 , , PUNCT bracis-33639 173 16 adjusting adjust VERB bracis-33639 173 17 the the DET bracis-33639 173 18 state state NOUN bracis-33639 173 19 vector vector NOUN bracis-33639 173 20 ’s ’s PART bracis-33639 173 21 size size NOUN bracis-33639 173 22 during during ADP bracis-33639 173 23 processing processing NOUN bracis-33639 173 24 . . PUNCT bracis-33639 174 1 table table NOUN bracis-33639 174 2     SPACE bracis-33639 174 3 2 2 NUM bracis-33639 174 4 provides provide VERB bracis-33639 174 5 the the DET bracis-33639 174 6 selected select VERB bracis-33639 174 7 hyperparameter hyperparameter NOUN bracis-33639 174 8 values value NOUN bracis-33639 174 9 . . PUNCT bracis-33639 175 1 baselines baseline NOUN bracis-33639 175 2 . . PUNCT bracis-33639 176 1 we we PRON bracis-33639 176 2 utilize utilize VERB bracis-33639 176 3 the the DET bracis-33639 176 4 default default NOUN bracis-33639 176 5 parameters parameter NOUN bracis-33639 176 6 provided provide VERB bracis-33639 176 7 by by ADP bracis-33639 176 8 the the DET bracis-33639 176 9 scikit scikit NOUN bracis-33639 176 10 - - PUNCT bracis-33639 176 11 learn learn VERB bracis-33639 176 12 python python NOUN bracis-33639 176 13 library library NOUN bracis-33639 176 14 [ [ X bracis-33639 176 15 23 23 NUM bracis-33639 176 16 ] ] PUNCT bracis-33639 176 17 for for ADP bracis-33639 176 18 all all DET bracis-33639 176 19 the the DET bracis-33639 176 20 baseline baseline ADJ bracis-33639 176 21 methods method NOUN bracis-33639 176 22 . . PUNCT bracis-33639 177 1 however however ADV bracis-33639 177 2 , , PUNCT bracis-33639 177 3 due due ADP bracis-33639 177 4 to to ADP bracis-33639 177 5 the the DET bracis-33639 177 6 dataset dataset NOUN bracis-33639 177 7 ’s ’s PART bracis-33639 177 8 imbalanced imbalanced ADJ bracis-33639 177 9 nature nature NOUN bracis-33639 177 10 , , PUNCT bracis-33639 177 11 we we PRON bracis-33639 177 12 employ employ VERB bracis-33639 177 13 the the DET bracis-33639 177 14 classweight classweight NOUN bracis-33639 177 15 parameter parameter NOUN bracis-33639 177 16 set set VERB bracis-33639 177 17 to to AUX bracis-33639 177 18 ’ ' PUNCT bracis-33639 177 19 balanced balanced ADJ bracis-33639 177 20 ’ ' PUNCT bracis-33639 177 21 across across ADP bracis-33639 177 22 all all DET bracis-33639 177 23 these these DET bracis-33639 177 24 methods method NOUN bracis-33639 177 25 . . PUNCT bracis-33639 178 1 this this DET bracis-33639 178 2 adjustment adjustment NOUN bracis-33639 178 3 ensures ensure VERB bracis-33639 178 4 more more ADV bracis-33639 178 5 accurate accurate ADJ bracis-33639 178 6 results result NOUN bracis-33639 178 7 by by ADP bracis-33639 178 8 addressing address VERB bracis-33639 178 9 the the DET bracis-33639 178 10 class class NOUN bracis-33639 178 11 imbalance imbalance NOUN bracis-33639 178 12 that that PRON bracis-33639 178 13 would would AUX bracis-33639 178 14 otherwise otherwise ADV bracis-33639 178 15 degrade degrade VERB bracis-33639 178 16 the the DET bracis-33639 178 17 model model NOUN bracis-33639 178 18 ’s ’s PART bracis-33639 178 19 performance performance NOUN bracis-33639 178 20 . . PUNCT bracis-33639 179 1 we we PRON bracis-33639 179 2 have have VERB bracis-33639 179 3 a a DET bracis-33639 179 4 list list NOUN bracis-33639 179 5 of of ADP bracis-33639 179 6 features feature NOUN bracis-33639 179 7 for for ADP bracis-33639 179 8 each each DET bracis-33639 179 9 time time NOUN bracis-33639 179 10 step step NOUN bracis-33639 179 11 , , PUNCT bracis-33639 179 12 with with ADP bracis-33639 179 13 each each DET bracis-33639 179 14 element element NOUN bracis-33639 179 15 representing represent VERB bracis-33639 179 16 a a DET bracis-33639 179 17 dataframe dataframe NOUN bracis-33639 179 18 of of ADP bracis-33639 179 19 node node NOUN bracis-33639 179 20 features feature NOUN bracis-33639 179 21 at at ADP bracis-33639 179 22 a a DET bracis-33639 179 23 specific specific ADJ bracis-33639 179 24 time time NOUN bracis-33639 179 25 step step NOUN bracis-33639 179 26 . . PUNCT bracis-33639 180 1 to to PART bracis-33639 180 2 prepare prepare VERB bracis-33639 180 3 the the DET bracis-33639 180 4 data datum NOUN bracis-33639 180 5 for for ADP bracis-33639 180 6 traditional traditional ADJ bracis-33639 180 7 baseline baseline ADJ bracis-33639 180 8 methods method NOUN bracis-33639 180 9 , , PUNCT bracis-33639 180 10 we we PRON bracis-33639 180 11 employed employ VERB bracis-33639 180 12 a a DET bracis-33639 180 13 sliding slide VERB bracis-33639 180 14 window window NOUN bracis-33639 180 15 approach approach NOUN bracis-33639 180 16 with with ADP bracis-33639 180 17 a a DET bracis-33639 180 18 window window NOUN bracis-33639 180 19 size size NOUN bracis-33639 180 20 of of ADP bracis-33639 180 21 3 3 NUM bracis-33639 180 22 . . PUNCT bracis-33639 181 1 in in ADP bracis-33639 181 2 this this DET bracis-33639 181 3 method method NOUN bracis-33639 181 4 , , PUNCT bracis-33639 181 5 we we PRON bracis-33639 181 6 selected select VERB bracis-33639 181 7 three three NUM bracis-33639 181 8 consecutive consecutive ADJ bracis-33639 181 9 dataframes dataframe NOUN bracis-33639 181 10 , , PUNCT bracis-33639 181 11 concatenated concatenate VERB bracis-33639 181 12 the the DET bracis-33639 181 13 first first ADJ bracis-33639 181 14 two two NUM bracis-33639 181 15 for for ADP bracis-33639 181 16 training training NOUN bracis-33639 181 17 , , PUNCT bracis-33639 181 18 and and CCONJ bracis-33639 181 19 used use VERB bracis-33639 181 20 the the DET bracis-33639 181 21 third third NOUN bracis-33639 181 22 for for ADP bracis-33639 181 23 testing testing NOUN bracis-33639 181 24 , , PUNCT bracis-33639 181 25 averaging average VERB bracis-33639 181 26 the the DET bracis-33639 181 27 results result NOUN bracis-33639 181 28 across across ADP bracis-33639 181 29 all all DET bracis-33639 181 30 sliding slide VERB bracis-33639 181 31 windows window NOUN bracis-33639 181 32 . . PUNCT bracis-33639 182 1 6 6 NUM bracis-33639 182 2 evaluation evaluation NOUN bracis-33639 182 3 we we PRON bracis-33639 182 4 used use VERB bracis-33639 182 5 precision precision NOUN bracis-33639 182 6 , , PUNCT bracis-33639 182 7 recall recall NOUN bracis-33639 182 8 , , PUNCT bracis-33639 182 9 and and CCONJ bracis-33639 182 10 f1 f1 NOUN bracis-33639 182 11 - - PUNCT bracis-33639 182 12 score score NOUN bracis-33639 182 13 to to PART bracis-33639 182 14 compare compare VERB bracis-33639 182 15 the the DET bracis-33639 182 16 models model NOUN bracis-33639 182 17 ’ ' PUNCT bracis-33639 182 18 performance performance NOUN bracis-33639 182 19 . . PUNCT bracis-33639 183 1 these these DET bracis-33639 183 2 metrics metric NOUN bracis-33639 183 3 are be AUX bracis-33639 183 4 calculated calculate VERB bracis-33639 183 5 for for ADP bracis-33639 183 6 both both CCONJ bracis-33639 183 7 the the DET bracis-33639 183 8 positive positive ADJ bracis-33639 183 9 class class NOUN bracis-33639 183 10 ( ( PUNCT bracis-33639 183 11 crime crime NOUN bracis-33639 183 12 occurrences occurrence NOUN bracis-33639 183 13 ) ) PUNCT bracis-33639 183 14 and and CCONJ bracis-33639 183 15 the the DET bracis-33639 183 16 negative negative ADJ bracis-33639 183 17 class class NOUN bracis-33639 183 18 ( ( PUNCT bracis-33639 183 19 no no DET bracis-33639 183 20 crime crime NOUN bracis-33639 183 21 occurrences occurrence NOUN bracis-33639 183 22 ) ) PUNCT bracis-33639 183 23 . . PUNCT bracis-33639 184 1 6.1 6.1 NUM bracis-33639 184 2 metrics metric NOUN bracis-33639 184 3 performance performance NOUN bracis-33639 184 4 in in ADP bracis-33639 184 5 both both DET bracis-33639 184 6 deep deep ADJ bracis-33639 184 7 learning learning NOUN bracis-33639 184 8 models model NOUN bracis-33639 184 9 , , PUNCT bracis-33639 184 10 as as SCONJ bracis-33639 184 11 illustrated illustrate VERB bracis-33639 184 12 in in ADP bracis-33639 184 13 fig fig NOUN bracis-33639 184 14 . . PUNCT bracis-33639 184 15     SPACE bracis-33639 185 1 2 2 NUM bracis-33639 185 2 , , PUNCT bracis-33639 185 3 the the DET bracis-33639 185 4 training training NOUN bracis-33639 185 5 and and CCONJ bracis-33639 185 6 validation validation NOUN bracis-33639 185 7 loss loss NOUN bracis-33639 185 8 curves curve NOUN bracis-33639 185 9 are be AUX bracis-33639 185 10 closely closely ADV bracis-33639 185 11 aligned align VERB bracis-33639 185 12 and and CCONJ bracis-33639 185 13 consistently consistently ADV bracis-33639 185 14 decrease decrease VERB bracis-33639 185 15 with with ADP bracis-33639 185 16 the the DET bracis-33639 185 17 increase increase NOUN bracis-33639 185 18 in in ADP bracis-33639 185 19 epochs epoch NOUN bracis-33639 185 20 . . PUNCT bracis-33639 186 1 the the DET bracis-33639 186 2 close close ADJ bracis-33639 186 3 alignment alignment NOUN bracis-33639 186 4 of of ADP bracis-33639 186 5 these these DET bracis-33639 186 6 curves curve NOUN bracis-33639 186 7 indicates indicate VERB bracis-33639 186 8 that that SCONJ bracis-33639 186 9 the the DET bracis-33639 186 10 models model NOUN bracis-33639 186 11 generalize generalize VERB bracis-33639 186 12 well well ADV bracis-33639 186 13 to to ADP bracis-33639 186 14 unseen unseen ADJ bracis-33639 186 15 data datum NOUN bracis-33639 186 16 , , PUNCT bracis-33639 186 17 providing provide VERB bracis-33639 186 18 strong strong ADJ bracis-33639 186 19 evidence evidence NOUN bracis-33639 186 20 that that SCONJ bracis-33639 186 21 overfitting overfitting NOUN bracis-33639 186 22 is be AUX bracis-33639 186 23 not not PART bracis-33639 186 24 occurring occur VERB bracis-33639 186 25 . . PUNCT bracis-33639 187 1 this this DET bracis-33639 187 2 observation observation NOUN bracis-33639 187 3 is be AUX bracis-33639 187 4 further far ADV bracis-33639 187 5 affirmed affirm VERB bracis-33639 187 6 by by ADP bracis-33639 187 7 comparing compare VERB bracis-33639 187 8 validation validation NOUN bracis-33639 187 9 and and CCONJ bracis-33639 187 10 test test NOUN bracis-33639 187 11 metrics metric NOUN bracis-33639 187 12 shown show VERB bracis-33639 187 13 in in ADP bracis-33639 187 14 fig fig NOUN bracis-33639 187 15 . . PUNCT bracis-33639 187 16     SPACE bracis-33639 188 1 3 3 X bracis-33639 188 2 . . PUNCT bracis-33639 189 1 however however ADV bracis-33639 189 2 , , PUNCT bracis-33639 189 3 during during ADP bracis-33639 189 4 the the DET bracis-33639 189 5 training training NOUN bracis-33639 189 6 of of ADP bracis-33639 189 7 the the DET bracis-33639 189 8 evolvegcn evolvegcn PROPN bracis-33639 189 9 model model NOUN bracis-33639 189 10 , , PUNCT bracis-33639 189 11 an an DET bracis-33639 189 12 unusual unusual ADJ bracis-33639 189 13 spike spike NOUN bracis-33639 189 14 occurs occur VERB bracis-33639 189 15 around around ADP bracis-33639 189 16 epoch epoch PROPN bracis-33639 189 17 800 800 NUM bracis-33639 189 18 . . PUNCT bracis-33639 190 1 we we PRON bracis-33639 190 2 hypothesize hypothesize VERB bracis-33639 190 3 that that SCONJ bracis-33639 190 4 this this PRON bracis-33639 190 5 may may AUX bracis-33639 190 6 be be AUX bracis-33639 190 7 due due ADJ bracis-33639 190 8 to to ADP bracis-33639 190 9 the the DET bracis-33639 190 10 model model NOUN bracis-33639 190 11 encountering encounter VERB bracis-33639 190 12 a a DET bracis-33639 190 13 local local ADJ bracis-33639 190 14 valley valley NOUN bracis-33639 190 15 in in ADP bracis-33639 190 16 the the DET bracis-33639 190 17 loss loss NOUN bracis-33639 190 18 landscape landscape NOUN bracis-33639 190 19 . . PUNCT bracis-33639 191 1 further further ADJ bracis-33639 191 2 investigation investigation NOUN bracis-33639 191 3 is be AUX bracis-33639 191 4 required require VERB bracis-33639 191 5 to to PART bracis-33639 191 6 confirm confirm VERB bracis-33639 191 7 this this DET bracis-33639 191 8 hypothesis hypothesis NOUN bracis-33639 191 9 . . PUNCT bracis-33639 192 1 fig fig NOUN bracis-33639 192 2 . . PUNCT bracis-33639 193 1 2 2 X bracis-33639 193 2 . . X bracis-33639 193 3 training training NOUN bracis-33639 193 4 and and CCONJ bracis-33639 193 5 validation validation NOUN bracis-33639 193 6 loss loss NOUN bracis-33639 193 7 over over ADP bracis-33639 193 8 epochs epoch NOUN bracis-33639 193 9 of of ADP bracis-33639 193 10 gnns gnns ADJ bracis-33639 193 11 models model NOUN bracis-33639 193 12 . . PUNCT bracis-33639 194 1 full full ADJ bracis-33639 194 2 size size NOUN bracis-33639 194 3 image image NOUN bracis-33639 194 4 dysat dysat VERB bracis-33639 194 5 . . PUNCT bracis-33639 195 1 in in ADP bracis-33639 195 2 the the DET bracis-33639 195 3 context context NOUN bracis-33639 195 4 of of ADP bracis-33639 195 5 highly highly ADV bracis-33639 195 6 imbalanced imbalanced ADJ bracis-33639 195 7 data datum NOUN bracis-33639 195 8 , , PUNCT bracis-33639 195 9 achieving achieve VERB bracis-33639 195 10 high high ADJ bracis-33639 195 11 f1 f1 NOUN bracis-33639 195 12 scores score NOUN bracis-33639 195 13 is be AUX bracis-33639 195 14 particularly particularly ADV bracis-33639 195 15 challenging challenging ADJ bracis-33639 195 16 but but CCONJ bracis-33639 195 17 essential essential ADJ bracis-33639 195 18 for for ADP bracis-33639 195 19 effective effective ADJ bracis-33639 195 20 model model NOUN bracis-33639 195 21 performance performance NOUN bracis-33639 195 22 . . PUNCT bracis-33639 196 1 in in ADP bracis-33639 196 2 fig fig NOUN bracis-33639 196 3 . . PUNCT bracis-33639 196 4     SPACE bracis-33639 196 5 3a 3a NUM bracis-33639 196 6 , , PUNCT bracis-33639 196 7 the the DET bracis-33639 196 8 increasing increase VERB bracis-33639 196 9 f1 f1 NOUN bracis-33639 196 10 scores score NOUN bracis-33639 196 11 for for ADP bracis-33639 196 12 both both CCONJ bracis-33639 196 13 validation validation NOUN bracis-33639 196 14 and and CCONJ bracis-33639 196 15 test test NOUN bracis-33639 196 16 sets set NOUN bracis-33639 196 17 indicate indicate VERB bracis-33639 196 18 that that SCONJ bracis-33639 196 19 the the DET bracis-33639 196 20 dysat dysat NOUN bracis-33639 196 21 model model NOUN bracis-33639 196 22 is be AUX bracis-33639 196 23 adept adept ADJ bracis-33639 196 24 at at ADP bracis-33639 196 25 handling handle VERB bracis-33639 196 26 the the DET bracis-33639 196 27 imbalanced imbalanced ADJ bracis-33639 196 28 dataset dataset NOUN bracis-33639 196 29 . . PUNCT bracis-33639 197 1 an an DET bracis-33639 197 2 f1 f1 ADJ bracis-33639 197 3 score score NOUN bracis-33639 197 4 of of ADP bracis-33639 197 5 0.57 0.57 NUM bracis-33639 197 6 on on ADP bracis-33639 197 7 the the DET bracis-33639 197 8 validation validation NOUN bracis-33639 197 9 set set NOUN bracis-33639 197 10 demonstrates demonstrate VERB bracis-33639 197 11 that that SCONJ bracis-33639 197 12 the the DET bracis-33639 197 13 model model NOUN bracis-33639 197 14 reasonably reasonably ADV bracis-33639 197 15 balances balance VERB bracis-33639 197 16 precision precision NOUN bracis-33639 197 17 and and CCONJ bracis-33639 197 18 recall recall NOUN bracis-33639 197 19 . . PUNCT bracis-33639 198 1 it it PRON bracis-33639 198 2 signifies signify VERB bracis-33639 198 3 that that PRON bracis-33639 198 4 dysat dysat NOUN bracis-33639 198 5 can can AUX bracis-33639 198 6 correctly correctly ADV bracis-33639 198 7 identify identify VERB bracis-33639 198 8 many many ADJ bracis-33639 198 9 positive positive ADJ bracis-33639 198 10 instances instance NOUN bracis-33639 198 11 while while SCONJ bracis-33639 198 12 minimizing minimize VERB bracis-33639 198 13 false false ADJ bracis-33639 198 14 positives positive NOUN bracis-33639 198 15 and and CCONJ bracis-33639 198 16 negatives negative NOUN bracis-33639 198 17 . . PUNCT bracis-33639 199 1 similarly similarly ADV bracis-33639 199 2 , , PUNCT bracis-33639 199 3 an an DET bracis-33639 199 4 f1 f1 ADJ bracis-33639 199 5 score score NOUN bracis-33639 199 6 of of ADP bracis-33639 199 7 around around ADP bracis-33639 199 8 0.55 0.55 NUM bracis-33639 199 9 on on ADP bracis-33639 199 10 the the DET bracis-33639 199 11 test test NOUN bracis-33639 199 12 set set NOUN bracis-33639 199 13 indicates indicate VERB bracis-33639 199 14 that that SCONJ bracis-33639 199 15 the the DET bracis-33639 199 16 model model NOUN bracis-33639 199 17 generalizes generalize VERB bracis-33639 199 18 well well ADV bracis-33639 199 19 to to ADP bracis-33639 199 20 unseen unseen ADJ bracis-33639 199 21 data datum NOUN bracis-33639 199 22 . . PUNCT bracis-33639 200 1 additionally additionally ADV bracis-33639 200 2 , , PUNCT bracis-33639 200 3 table table NOUN bracis-33639 200 4     SPACE bracis-33639 200 5 3 3 NUM bracis-33639 200 6 reaffirms reaffirm VERB bracis-33639 200 7 dysat dysat ADJ bracis-33639 200 8 ’s ’s PART bracis-33639 200 9 superior superior ADJ bracis-33639 200 10 performance performance NOUN bracis-33639 200 11 compared compare VERB bracis-33639 200 12 to to ADP bracis-33639 200 13 other other ADJ bracis-33639 200 14 methods method NOUN bracis-33639 200 15 . . PUNCT bracis-33639 201 1 the the DET bracis-33639 201 2 dysat dysat NOUN bracis-33639 201 3 model model NOUN bracis-33639 201 4 ’s ’s PART bracis-33639 201 5 ability ability NOUN bracis-33639 201 6 to to PART bracis-33639 201 7 manage manage VERB bracis-33639 201 8 the the DET bracis-33639 201 9 complexities complexity NOUN bracis-33639 201 10 of of ADP bracis-33639 201 11 imbalanced imbalanced ADJ bracis-33639 201 12 data datum NOUN bracis-33639 201 13 and and CCONJ bracis-33639 201 14 maintain maintain VERB bracis-33639 201 15 consistent consistent ADJ bracis-33639 201 16 performance performance NOUN bracis-33639 201 17 across across ADP bracis-33639 201 18 validation validation NOUN bracis-33639 201 19 and and CCONJ bracis-33639 201 20 test test NOUN bracis-33639 201 21 sets set NOUN bracis-33639 201 22 highlights highlight VERB bracis-33639 201 23 its its PRON bracis-33639 201 24 potential potential NOUN bracis-33639 201 25 for for ADP bracis-33639 201 26 real real ADJ bracis-33639 201 27 - - PUNCT bracis-33639 201 28 world world NOUN bracis-33639 201 29 applications application NOUN bracis-33639 201 30 , , PUNCT bracis-33639 201 31 particularly particularly ADV bracis-33639 201 32 in in ADP bracis-33639 201 33 areas area NOUN bracis-33639 201 34 where where SCONJ bracis-33639 201 35 accurate accurate ADJ bracis-33639 201 36 minority minority NOUN bracis-33639 201 37 class class NOUN bracis-33639 201 38 predictions prediction NOUN bracis-33639 201 39 are be AUX bracis-33639 201 40 critical critical ADJ bracis-33639 201 41 . . PUNCT bracis-33639 202 1 fig fig NOUN bracis-33639 202 2 . . PUNCT bracis-33639 203 1 3 3 X bracis-33639 203 2 . . X bracis-33639 203 3 f1 f1 PROPN bracis-33639 203 4 scores score NOUN bracis-33639 203 5 over over ADP bracis-33639 203 6 epochs epoch NOUN bracis-33639 203 7 of of ADP bracis-33639 203 8 gnn gnn PROPN bracis-33639 203 9 ’s ’s NOUN bracis-33639 203 10     SPACE bracis-33639 203 11 model model NOUN bracis-33639 203 12 full full ADJ bracis-33639 203 13 size size NOUN bracis-33639 203 14 image image NOUN bracis-33639 203 15 evolvegcn evolvegcn ADV bracis-33639 203 16 . . PUNCT bracis-33639 204 1 despite despite SCONJ bracis-33639 204 2 the the DET bracis-33639 204 3 evolvegcn evolvegcn PROPN bracis-33639 204 4 model model PROPN bracis-33639 204 5 ’s ’s PART bracis-33639 204 6 metrics metric NOUN bracis-33639 204 7 not not PART bracis-33639 204 8 ranking rank VERB bracis-33639 204 9 among among ADP bracis-33639 204 10 the the DET bracis-33639 204 11 highest high ADJ bracis-33639 204 12 , , PUNCT bracis-33639 204 13 when when SCONJ bracis-33639 204 14 we we PRON bracis-33639 204 15 plot plot VERB bracis-33639 204 16 the the DET bracis-33639 204 17 model model NOUN bracis-33639 204 18 ’s ’s PART bracis-33639 204 19 results result NOUN bracis-33639 204 20 on on ADP bracis-33639 204 21 the the DET bracis-33639 204 22 map map NOUN bracis-33639 204 23 for for ADP bracis-33639 204 24 hotspot hotspot NOUN bracis-33639 204 25 analysis analysis NOUN bracis-33639 204 26 , , PUNCT bracis-33639 204 27 we we PRON bracis-33639 204 28 observe observe VERB bracis-33639 204 29 an an DET bracis-33639 204 30 interesting interesting ADJ bracis-33639 204 31 outcome outcome NOUN bracis-33639 204 32 . . PUNCT bracis-33639 205 1 as as SCONJ bracis-33639 205 2 illustrated illustrate VERB bracis-33639 205 3 in in ADP bracis-33639 205 4 fig fig NOUN bracis-33639 205 5 . . PUNCT bracis-33639 205 6     SPACE bracis-33639 206 1 4 4 NUM bracis-33639 206 2 , , PUNCT bracis-33639 206 3 the the DET bracis-33639 206 4 reason reason NOUN bracis-33639 206 5 for for ADP bracis-33639 206 6 the the DET bracis-33639 206 7 metrics metric NOUN bracis-33639 206 8 not not PART bracis-33639 206 9 being be AUX bracis-33639 206 10 so so ADV bracis-33639 206 11 high high ADJ bracis-33639 206 12 is be AUX bracis-33639 206 13 that that SCONJ bracis-33639 206 14 the the DET bracis-33639 206 15 model model NOUN bracis-33639 206 16 tends tend VERB bracis-33639 206 17 to to PART bracis-33639 206 18 favor favor VERB bracis-33639 206 19 classifying classify VERB bracis-33639 206 20 safe safe ADJ bracis-33639 206 21 ( ( PUNCT bracis-33639 206 22 blue blue ADJ bracis-33639 206 23 ) ) PUNCT bracis-33639 206 24 and and CCONJ bracis-33639 206 25 unsafe unsafe ADJ bracis-33639 206 26 ( ( PUNCT bracis-33639 206 27 red red ADJ bracis-33639 206 28 ) ) PUNCT bracis-33639 206 29 areas area NOUN bracis-33639 206 30 only only ADV bracis-33639 206 31 when when SCONJ bracis-33639 206 32 there there PRON bracis-33639 206 33 is be VERB bracis-33639 206 34 strong strong ADJ bracis-33639 206 35 evidence evidence NOUN bracis-33639 206 36 to to PART bracis-33639 206 37 do do AUX bracis-33639 206 38 so so ADV bracis-33639 206 39 , , PUNCT bracis-33639 206 40 which which PRON bracis-33639 206 41 explains explain VERB bracis-33639 206 42 the the DET bracis-33639 206 43 large large ADJ bracis-33639 206 44 number number NOUN bracis-33639 206 45 of of ADP bracis-33639 206 46 uncertain uncertain ADJ bracis-33639 206 47 areas area NOUN bracis-33639 206 48 ( ( PUNCT bracis-33639 206 49 blue blue ADJ bracis-33639 206 50 ) ) PUNCT bracis-33639 206 51 and and CCONJ bracis-33639 206 52 the the DET bracis-33639 206 53 higher high ADJ bracis-33639 206 54 recall recall NOUN bracis-33639 206 55 compared compare VERB bracis-33639 206 56 to to ADP bracis-33639 206 57 all all DET bracis-33639 206 58 other other ADJ bracis-33639 206 59 models model NOUN bracis-33639 206 60 . . PUNCT bracis-33639 207 1 this this DET bracis-33639 207 2 aspect aspect NOUN bracis-33639 207 3 of of ADP bracis-33639 207 4 the the DET bracis-33639 207 5 model model NOUN bracis-33639 207 6 has have VERB bracis-33639 207 7 its its PRON bracis-33639 207 8 advantages advantage NOUN bracis-33639 207 9 and and CCONJ bracis-33639 207 10 disadvantages disadvantage NOUN bracis-33639 207 11 . . PUNCT bracis-33639 208 1 for for ADP bracis-33639 208 2 instance instance NOUN bracis-33639 208 3 , , PUNCT bracis-33639 208 4 while while SCONJ bracis-33639 208 5 this this DET bracis-33639 208 6 cautious cautious ADJ bracis-33639 208 7 approach approach NOUN bracis-33639 208 8 may may AUX bracis-33639 208 9 lead lead VERB bracis-33639 208 10 to to ADP bracis-33639 208 11 more more ADV bracis-33639 208 12 reliable reliable ADJ bracis-33639 208 13 identifications identification NOUN bracis-33639 208 14 of of ADP bracis-33639 208 15 true true ADJ bracis-33639 208 16 hotspots hotspot NOUN bracis-33639 208 17 , , PUNCT bracis-33639 208 18 it it PRON bracis-33639 208 19 can can AUX bracis-33639 208 20 also also ADV bracis-33639 208 21 result result VERB bracis-33639 208 22 in in ADP bracis-33639 208 23 overlooking overlook VERB bracis-33639 208 24 subtle subtle ADJ bracis-33639 208 25 but but CCONJ bracis-33639 208 26 important important ADJ bracis-33639 208 27 patterns pattern NOUN bracis-33639 208 28 or or CCONJ bracis-33639 208 29 changes change NOUN bracis-33639 208 30 in in ADP bracis-33639 208 31 the the DET bracis-33639 208 32 data datum NOUN bracis-33639 208 33 , , PUNCT bracis-33639 208 34 thereby thereby ADV bracis-33639 208 35 affecting affect VERB bracis-33639 208 36 the the DET bracis-33639 208 37 model model NOUN bracis-33639 208 38 ’s ’s PART bracis-33639 208 39 overall overall ADJ bracis-33639 208 40 precision precision NOUN bracis-33639 208 41 and and CCONJ bracis-33639 208 42 specificity specificity NOUN bracis-33639 208 43 . . PUNCT bracis-33639 209 1 table table NOUN bracis-33639 209 2 3 3 NUM bracis-33639 209 3 . . PUNCT bracis-33639 209 4 performance performance NOUN bracis-33639 209 5 metrics.full metrics.full NUM bracis-33639 209 6 size size NOUN bracis-33639 209 7 table table NOUN bracis-33639 209 8 fig fig NOUN bracis-33639 209 9 . . PUNCT bracis-33639 210 1 4 4 X bracis-33639 210 2 . . PUNCT bracis-33639 210 3 model model NOUN bracis-33639 210 4 predictions prediction NOUN bracis-33639 210 5 projected project VERB bracis-33639 210 6 onto onto ADP bracis-33639 210 7 street street NOUN bracis-33639 210 8 graph graph NOUN bracis-33639 210 9 . . PUNCT bracis-33639 211 1 ( ( PUNCT bracis-33639 211 2 color color NOUN bracis-33639 211 3 figure figure NOUN bracis-33639 211 4 online online ADV bracis-33639 211 5 ) ) PUNCT bracis-33639 211 6 full full ADJ bracis-33639 211 7 size size NOUN bracis-33639 211 8 image image NOUN bracis-33639 211 9 baselines baseline NOUN bracis-33639 211 10 . . PUNCT bracis-33639 212 1 when when SCONJ bracis-33639 212 2 it it PRON bracis-33639 212 3 comes come VERB bracis-33639 212 4 to to ADP bracis-33639 212 5 traditional traditional ADJ bracis-33639 212 6 machine machine NOUN bracis-33639 212 7 learning learning NOUN bracis-33639 212 8 models model NOUN bracis-33639 212 9 , , PUNCT bracis-33639 212 10 random random ADJ bracis-33639 212 11 forest forest NOUN bracis-33639 212 12 achieved achieve VERB bracis-33639 212 13 results result NOUN bracis-33639 212 14 comparable comparable ADJ bracis-33639 212 15 to to ADP bracis-33639 212 16 the the DET bracis-33639 212 17 dysat dysat NOUN bracis-33639 212 18 model model NOUN bracis-33639 212 19 for for ADP bracis-33639 212 20 class class NOUN bracis-33639 212 21 1 1 NUM bracis-33639 212 22 metrics metric NOUN bracis-33639 212 23 , , PUNCT bracis-33639 212 24 while while SCONJ bracis-33639 212 25 logistic logistic ADJ bracis-33639 212 26 regression regression NOUN bracis-33639 212 27 performed perform VERB bracis-33639 212 28 similarly similarly ADV bracis-33639 212 29 to to ADP bracis-33639 212 30 the the DET bracis-33639 212 31 evolvegcn evolvegcn PROPN bracis-33639 212 32 model model NOUN bracis-33639 212 33 , , PUNCT bracis-33639 212 34 as as SCONJ bracis-33639 212 35 shown show VERB bracis-33639 212 36 in in ADP bracis-33639 212 37 table table NOUN bracis-33639 212 38     SPACE bracis-33639 212 39 3 3 NUM bracis-33639 212 40 . . PUNCT bracis-33639 213 1 in in ADP bracis-33639 213 2 contrast contrast NOUN bracis-33639 213 3 , , PUNCT bracis-33639 213 4 the the DET bracis-33639 213 5 svm svm ADJ bracis-33639 213 6 model model NOUN bracis-33639 213 7 demonstrated demonstrate VERB bracis-33639 213 8 notably notably ADV bracis-33639 213 9 poor poor ADJ bracis-33639 213 10 performance performance NOUN bracis-33639 213 11 . . PUNCT bracis-33639 214 1 despite despite SCONJ bracis-33639 214 2 the the DET bracis-33639 214 3 metrics metric NOUN bracis-33639 214 4 being be AUX bracis-33639 214 5 similar similar ADJ bracis-33639 214 6 to to ADP bracis-33639 214 7 those those PRON bracis-33639 214 8 of of ADP bracis-33639 214 9 graph graph NOUN bracis-33639 214 10 - - PUNCT bracis-33639 214 11 based base VERB bracis-33639 214 12 models model NOUN bracis-33639 214 13 , , PUNCT bracis-33639 214 14 there there PRON bracis-33639 214 15 is be VERB bracis-33639 214 16 a a DET bracis-33639 214 17 crucial crucial ADJ bracis-33639 214 18 caveat caveat NOUN bracis-33639 214 19 to to PART bracis-33639 214 20 consider consider VERB bracis-33639 214 21 : : PUNCT bracis-33639 214 22 traditional traditional ADJ bracis-33639 214 23 models model NOUN bracis-33639 214 24 do do AUX bracis-33639 214 25 not not PART bracis-33639 214 26 take take VERB bracis-33639 214 27 into into ADP bracis-33639 214 28 account account NOUN bracis-33639 214 29 the the DET bracis-33639 214 30 temporal temporal ADJ bracis-33639 214 31 dynamics dynamic NOUN bracis-33639 214 32 of of ADP bracis-33639 214 33 our our PRON bracis-33639 214 34 data datum NOUN bracis-33639 214 35 . . PUNCT bracis-33639 215 1 consequently consequently ADV bracis-33639 215 2 , , PUNCT bracis-33639 215 3 they they PRON bracis-33639 215 4 fail fail VERB bracis-33639 215 5 to to PART bracis-33639 215 6 generalize generalize VERB bracis-33639 215 7 their their PRON bracis-33639 215 8 predictions prediction NOUN bracis-33639 215 9 over over ADP bracis-33639 215 10 time time NOUN bracis-33639 215 11 . . PUNCT bracis-33639 216 1 this this DET bracis-33639 216 2 limitation limitation NOUN bracis-33639 216 3 becomes become VERB bracis-33639 216 4 particularly particularly ADV bracis-33639 216 5 evident evident ADJ bracis-33639 216 6 when when SCONJ bracis-33639 216 7 plotting plot VERB bracis-33639 216 8 predictions prediction NOUN bracis-33639 216 9 on on ADP bracis-33639 216 10 a a DET bracis-33639 216 11 map map NOUN bracis-33639 216 12 for for ADP bracis-33639 216 13 hotspot hotspot NOUN bracis-33639 216 14 construction construction NOUN bracis-33639 216 15 , , PUNCT bracis-33639 216 16 as as SCONJ bracis-33639 216 17 we we PRON bracis-33639 216 18 show show VERB bracis-33639 216 19 in in ADP bracis-33639 216 20 fig fig NOUN bracis-33639 216 21 . . PUNCT bracis-33639 216 22     SPACE bracis-33639 217 1 4 4 NUM bracis-33639 217 2 . . PUNCT bracis-33639 217 3 in in ADP bracis-33639 217 4 these these DET bracis-33639 217 5 figures figure NOUN bracis-33639 217 6 , , PUNCT bracis-33639 217 7 green green PROPN bracis-33639 217 8 represents represent VERB bracis-33639 217 9 “ " PUNCT bracis-33639 217 10 safe safe ADJ bracis-33639 217 11 ” " PUNCT bracis-33639 217 12 zones zone NOUN bracis-33639 217 13 , , PUNCT bracis-33639 217 14 red red PROPN bracis-33639 217 15 indicates indicate VERB bracis-33639 217 16 “ " PUNCT bracis-33639 217 17 dangerous dangerous ADJ bracis-33639 217 18 ” " PUNCT bracis-33639 217 19 zones zone NOUN bracis-33639 217 20 , , PUNCT bracis-33639 217 21 and and CCONJ bracis-33639 217 22 blue blue ADJ bracis-33639 217 23 signifies signifie NOUN bracis-33639 217 24 model model NOUN bracis-33639 217 25 uncertainty uncertainty NOUN bracis-33639 217 26 . . PUNCT bracis-33639 218 1 notably notably ADV bracis-33639 218 2 , , PUNCT bracis-33639 218 3 the the DET bracis-33639 218 4 gnn gnn PROPN bracis-33639 218 5 models model NOUN bracis-33639 218 6 have have VERB bracis-33639 218 7 distinct distinct ADJ bracis-33639 218 8 safe safe ADJ bracis-33639 218 9 and and CCONJ bracis-33639 218 10 dangerous dangerous ADJ bracis-33639 218 11 regions region NOUN bracis-33639 218 12 , , PUNCT bracis-33639 218 13 which which PRON bracis-33639 218 14 we we PRON bracis-33639 218 15 will will AUX bracis-33639 218 16 discuss discuss VERB bracis-33639 218 17 in in ADP bracis-33639 218 18 the the DET bracis-33639 218 19 next next ADJ bracis-33639 218 20 section section NOUN bracis-33639 218 21 . . PUNCT bracis-33639 219 1 however however ADV bracis-33639 219 2 , , PUNCT bracis-33639 219 3 in in ADP bracis-33639 219 4 traditional traditional ADJ bracis-33639 219 5 models model NOUN bracis-33639 219 6 , , PUNCT bracis-33639 219 7 these these DET bracis-33639 219 8 regions region NOUN bracis-33639 219 9 are be AUX bracis-33639 219 10 less less ADV bracis-33639 219 11 clear clear ADJ bracis-33639 219 12 . . PUNCT bracis-33639 220 1 this this PRON bracis-33639 220 2 is be AUX bracis-33639 220 3 because because SCONJ bracis-33639 220 4 the the DET bracis-33639 220 5 models model NOUN bracis-33639 220 6 are be AUX bracis-33639 220 7 often often ADV bracis-33639 220 8 uncertain uncertain ADJ bracis-33639 220 9 about about ADP bracis-33639 220 10 their their PRON bracis-33639 220 11 predictions prediction NOUN bracis-33639 220 12 ( ( PUNCT bracis-33639 220 13 as as SCONJ bracis-33639 220 14 seen see VERB bracis-33639 220 15 with with ADP bracis-33639 220 16 svm svm ADJ bracis-33639 220 17 , , PUNCT bracis-33639 220 18 logistic logistic ADJ bracis-33639 220 19 regression regression NOUN bracis-33639 220 20 and and CCONJ bracis-33639 220 21 even even ADV bracis-33639 220 22 random random ADJ bracis-33639 220 23 forest forest NOUN bracis-33639 220 24 have have VERB bracis-33639 220 25 more more ADJ bracis-33639 220 26 blue blue ADJ bracis-33639 220 27 lines line NOUN bracis-33639 220 28 ) ) PUNCT bracis-33639 220 29 or or CCONJ bracis-33639 220 30 because because SCONJ bracis-33639 220 31 they they PRON bracis-33639 220 32 make make VERB bracis-33639 220 33 highly highly ADV bracis-33639 220 34 heterogeneous heterogeneous ADJ bracis-33639 220 35 predictions prediction NOUN bracis-33639 220 36 across across ADP bracis-33639 220 37 the the DET bracis-33639 220 38 region region NOUN bracis-33639 220 39 of of ADP bracis-33639 220 40 interest interest NOUN bracis-33639 220 41 . . PUNCT bracis-33639 221 1 6.2 6.2 NUM bracis-33639 221 2 hotspot hotspot NOUN bracis-33639 221 3 analysis analysis NOUN bracis-33639 221 4 as as SCONJ bracis-33639 221 5 shown show VERB bracis-33639 221 6 in in ADP bracis-33639 221 7 fig fig NOUN bracis-33639 221 8 . . PUNCT bracis-33639 221 9     SPACE bracis-33639 221 10 5 5 NUM bracis-33639 221 11 , , PUNCT bracis-33639 221 12 predictions prediction NOUN bracis-33639 221 13 produced produce VERB bracis-33639 221 14 by by ADP bracis-33639 221 15 gnn gnn PROPN bracis-33639 221 16 methods method NOUN bracis-33639 221 17 , , PUNCT bracis-33639 221 18 which which PRON bracis-33639 221 19 consider consider VERB bracis-33639 221 20 the the DET bracis-33639 221 21 temporal temporal ADJ bracis-33639 221 22 dynamics dynamic NOUN bracis-33639 221 23 of of ADP bracis-33639 221 24 the the DET bracis-33639 221 25 events event NOUN bracis-33639 221 26 under under ADP bracis-33639 221 27 study study NOUN bracis-33639 221 28 , , PUNCT bracis-33639 221 29 can can AUX bracis-33639 221 30 pinpoint pinpoint VERB bracis-33639 221 31 several several ADJ bracis-33639 221 32 notable notable ADJ bracis-33639 221 33 regions region NOUN bracis-33639 221 34 on on ADP bracis-33639 221 35 the the DET bracis-33639 221 36 map map NOUN bracis-33639 221 37 , , PUNCT bracis-33639 221 38 marked mark VERB bracis-33639 221 39 a a DET bracis-33639 221 40 , , PUNCT bracis-33639 221 41 b b NOUN bracis-33639 221 42 , , PUNCT bracis-33639 221 43 c c NOUN bracis-33639 221 44 , , PUNCT bracis-33639 221 45 d d NOUN bracis-33639 221 46 , , PUNCT bracis-33639 221 47 e e NOUN bracis-33639 221 48 , , PUNCT bracis-33639 221 49 and and CCONJ bracis-33639 221 50 f. f. NOUN bracis-33639 221 51 these these DET bracis-33639 221 52 regions region NOUN bracis-33639 221 53 are be AUX bracis-33639 221 54 significant significant ADJ bracis-33639 221 55 as as SCONJ bracis-33639 221 56 they they PRON bracis-33639 221 57 align align VERB bracis-33639 221 58 with with ADP bracis-33639 221 59 known know VERB bracis-33639 221 60 areas area NOUN bracis-33639 221 61 of of ADP bracis-33639 221 62 high high ADJ bracis-33639 221 63 or or CCONJ bracis-33639 221 64 low low ADJ bracis-33639 221 65 crime crime NOUN bracis-33639 221 66 rates rate NOUN bracis-33639 221 67 in in ADP bracis-33639 221 68 central central ADJ bracis-33639 221 69 são são PROPN bracis-33639 221 70 paulo paulo PROPN bracis-33639 221 71 , , PUNCT bracis-33639 221 72 as as SCONJ bracis-33639 221 73 shown show VERB bracis-33639 221 74 in in ADP bracis-33639 221 75 fig fig NOUN bracis-33639 221 76 . . PUNCT bracis-33639 221 77     SPACE bracis-33639 222 1 1 1 NUM bracis-33639 222 2 . . PUNCT bracis-33639 223 1 a a DET bracis-33639 223 2 detailed detailed ADJ bracis-33639 223 3 description description NOUN bracis-33639 223 4 of of ADP bracis-33639 223 5 each each DET bracis-33639 223 6 region region NOUN bracis-33639 223 7 is be AUX bracis-33639 223 8 provided provide VERB bracis-33639 223 9 below below ADP bracis-33639 223 10 : : PUNCT bracis-33639 223 11 a. a. NOUN bracis-33639 223 12 encompasses encompass VERB bracis-33639 223 13 the the DET bracis-33639 223 14 entire entire ADJ bracis-33639 223 15 república república NOUN bracis-33639 223 16 neighborhood neighborhood NOUN bracis-33639 223 17 in in ADP bracis-33639 223 18 são são PROPN bracis-33639 223 19 paulo paulo PROPN bracis-33639 223 20 , , PUNCT bracis-33639 223 21 which which PRON bracis-33639 223 22 is be AUX bracis-33639 223 23 widely widely ADV bracis-33639 223 24 known know VERB bracis-33639 223 25 for for ADP bracis-33639 223 26 its its PRON bracis-33639 223 27 high high ADJ bracis-33639 223 28 risk risk NOUN bracis-33639 223 29 of of ADP bracis-33639 223 30 robbery robbery NOUN bracis-33639 223 31 and and CCONJ bracis-33639 223 32 theft theft NOUN bracis-33639 223 33 . . PUNCT bracis-33639 224 1 b. b. PROPN bracis-33639 224 2 located locate VERB bracis-33639 224 3 just just ADV bracis-33639 224 4 above above ADP bracis-33639 224 5 the the DET bracis-33639 224 6 república república PROPN bracis-33639 224 7 neighborhood neighborhood NOUN bracis-33639 224 8 , , PUNCT bracis-33639 224 9 is be AUX bracis-33639 224 10 identified identify VERB bracis-33639 224 11 by by ADP bracis-33639 224 12 our our PRON bracis-33639 224 13 model model NOUN bracis-33639 224 14 as as ADP bracis-33639 224 15 a a DET bracis-33639 224 16 safer safe ADJ bracis-33639 224 17 area area NOUN bracis-33639 224 18 . . PUNCT bracis-33639 225 1 examination examination NOUN bracis-33639 225 2 of of ADP bracis-33639 225 3 this this DET bracis-33639 225 4 region region NOUN bracis-33639 225 5 reveals reveal VERB bracis-33639 225 6 the the DET bracis-33639 225 7 presence presence NOUN bracis-33639 225 8 of of ADP bracis-33639 225 9 several several ADJ bracis-33639 225 10 banks bank NOUN bracis-33639 225 11 , , PUNCT bracis-33639 225 12 including include VERB bracis-33639 225 13 bradesco bradesco NOUN bracis-33639 225 14 , , PUNCT bracis-33639 225 15 santander santander NOUN bracis-33639 225 16 , , PUNCT bracis-33639 225 17 and and CCONJ bracis-33639 225 18 itaú itaú PROPN bracis-33639 225 19 , , PUNCT bracis-33639 225 20 in in ADP bracis-33639 225 21 the the DET bracis-33639 225 22 nearby nearby ADJ bracis-33639 225 23 blocks block NOUN bracis-33639 225 24 . . PUNCT bracis-33639 226 1 c. c. NOUN bracis-33639 226 2 includes include VERB bracis-33639 226 3 the the DET bracis-33639 226 4 surroundings surrounding NOUN bracis-33639 226 5 of of ADP bracis-33639 226 6 the the DET bracis-33639 226 7 estação estação PROPN bracis-33639 226 8 da da PROPN bracis-33639 226 9 luz luz PROPN bracis-33639 226 10 in in ADP bracis-33639 226 11 são são PROPN bracis-33639 226 12 paulo paulo PROPN bracis-33639 226 13 , , PUNCT bracis-33639 226 14 a a DET bracis-33639 226 15 well well ADV bracis-33639 226 16 - - PUNCT bracis-33639 226 17 known know VERB bracis-33639 226 18 high high ADJ bracis-33639 226 19 - - PUNCT bracis-33639 226 20 traffic traffic NOUN bracis-33639 226 21 area area NOUN bracis-33639 226 22 with with ADP bracis-33639 226 23 a a DET bracis-33639 226 24 reputation reputation NOUN bracis-33639 226 25 for for ADP bracis-33639 226 26 being be AUX bracis-33639 226 27 dangerous dangerous ADJ bracis-33639 226 28 . . PUNCT bracis-33639 227 1 d. d. PROPN bracis-33639 227 2 covers cover VERB bracis-33639 227 3 the the DET bracis-33639 227 4 vicinity vicinity NOUN bracis-33639 227 5 of of ADP bracis-33639 227 6 largo largo PROPN bracis-33639 227 7 da da PROPN bracis-33639 227 8 concórdia concórdia PROPN bracis-33639 227 9 square square PROPN bracis-33639 227 10 in in ADP bracis-33639 227 11 são são PROPN bracis-33639 227 12 paulo paulo PROPN bracis-33639 227 13 , , PUNCT bracis-33639 227 14 near near ADP bracis-33639 227 15 a a DET bracis-33639 227 16 lojão lojão NOUN bracis-33639 227 17 do do AUX bracis-33639 227 18 brás brás NOUN bracis-33639 227 19 . . PUNCT bracis-33639 228 1 this this DET bracis-33639 228 2 area area NOUN bracis-33639 228 3 , , PUNCT bracis-33639 228 4 where where SCONJ bracis-33639 228 5 most most ADJ bracis-33639 228 6 transportation transportation NOUN bracis-33639 228 7 is be AUX bracis-33639 228 8 on on ADP bracis-33639 228 9 foot foot NOUN bracis-33639 228 10 , , PUNCT bracis-33639 228 11 experiences experience VERB bracis-33639 228 12 a a DET bracis-33639 228 13 high high ADJ bracis-33639 228 14 incidence incidence NOUN bracis-33639 228 15 of of ADP bracis-33639 228 16 pickpocketing pickpockete VERB bracis-33639 228 17 , , PUNCT bracis-33639 228 18 leading lead VERB bracis-33639 228 19 to to ADP bracis-33639 228 20 many many ADJ bracis-33639 228 21 cell cell NOUN bracis-33639 228 22 phone phone NOUN bracis-33639 228 23 thefts theft NOUN bracis-33639 228 24 . . PUNCT bracis-33639 229 1 e. e. PROPN bracis-33639 229 2 encompasses encompass VERB bracis-33639 229 3 regions region NOUN bracis-33639 229 4 a a PRON bracis-33639 229 5 and and CCONJ bracis-33639 229 6 is be AUX bracis-33639 229 7 very very ADV bracis-33639 229 8 close close ADJ bracis-33639 229 9 to to ADP bracis-33639 229 10 region region PROPN bracis-33639 229 11 c. c. PROPN bracis-33639 229 12 f. f. PROPN bracis-33639 229 13 encompasses encompass VERB bracis-33639 229 14 region region PROPN bracis-33639 229 15 d. d. PROPN bracis-33639 229 16 fig fig PROPN bracis-33639 229 17 . . PUNCT bracis-33639 230 1 5 5 X bracis-33639 230 2 . . X bracis-33639 230 3 noteworthy noteworthy ADJ bracis-33639 230 4 predicted predict VERB bracis-33639 230 5 regions region NOUN bracis-33639 230 6 . . PUNCT bracis-33639 231 1 ( ( PUNCT bracis-33639 231 2 color color NOUN bracis-33639 231 3 figure figure NOUN bracis-33639 231 4 online online ADV bracis-33639 231 5 ) ) PUNCT bracis-33639 231 6 full full ADJ bracis-33639 231 7 size size NOUN bracis-33639 231 8 image image NOUN bracis-33639 231 9 in in ADP bracis-33639 231 10 our our PRON bracis-33639 231 11 hotspot hotspot NOUN bracis-33639 231 12 maps map NOUN bracis-33639 231 13 , , PUNCT bracis-33639 231 14 shown show VERB bracis-33639 231 15 in in ADP bracis-33639 231 16 figs fig NOUN bracis-33639 231 17 . . PUNCT bracis-33639 231 18     SPACE bracis-33639 231 19 4 4 NUM bracis-33639 231 20 and and CCONJ bracis-33639 231 21 5 5 NUM bracis-33639 231 22 , , PUNCT bracis-33639 231 23 we we PRON bracis-33639 231 24 used use VERB bracis-33639 231 25 soft soft ADJ bracis-33639 231 26 predictions prediction NOUN bracis-33639 231 27 for for ADP bracis-33639 231 28 both both CCONJ bracis-33639 231 29 baseline baseline ADJ bracis-33639 231 30 and and CCONJ bracis-33639 231 31 gnn gnn PROPN bracis-33639 231 32 models model NOUN bracis-33639 231 33 to to PART bracis-33639 231 34 better well ADV bracis-33639 231 35 represent represent VERB bracis-33639 231 36 prediction prediction NOUN bracis-33639 231 37 confidence confidence NOUN bracis-33639 231 38 . . PUNCT bracis-33639 232 1 probabilities probability NOUN bracis-33639 232 2 were be AUX bracis-33639 232 3 mapped map VERB bracis-33639 232 4 into into ADP bracis-33639 232 5 three three NUM bracis-33639 232 6 regions region NOUN bracis-33639 232 7 : : PUNCT bracis-33639 232 8 non non ADJ bracis-33639 232 9 - - NOUN bracis-33639 232 10 crime crime ADJ bracis-33639 232 11 ( ( PUNCT bracis-33639 232 12 green green NOUN bracis-33639 232 13 ) ) PUNCT bracis-33639 232 14 for for ADP bracis-33639 232 15 probabilities probability NOUN bracis-33639 232 16 near near ADP bracis-33639 232 17 0 0 NUM bracis-33639 232 18 , , PUNCT bracis-33639 232 19 uncertain uncertain ADJ bracis-33639 232 20 ( ( PUNCT bracis-33639 232 21 blue blue NOUN bracis-33639 232 22 ) ) PUNCT bracis-33639 232 23 for for ADP bracis-33639 232 24 probabilities probability NOUN bracis-33639 232 25 around around ADP bracis-33639 232 26 0.5 0.5 NUM bracis-33639 232 27 , , PUNCT bracis-33639 232 28 and and CCONJ bracis-33639 232 29 crime crime NOUN bracis-33639 232 30 ( ( PUNCT bracis-33639 232 31 red red PROPN bracis-33639 232 32 ) ) PUNCT bracis-33639 232 33 for for ADP bracis-33639 232 34 probabilities probability NOUN bracis-33639 232 35 near near ADP bracis-33639 232 36 1 1 NUM bracis-33639 232 37 . . PUNCT bracis-33639 233 1 this this DET bracis-33639 233 2 mapping mapping NOUN bracis-33639 233 3 effectively effectively ADV bracis-33639 233 4 highlights highlight VERB bracis-33639 233 5 areas area NOUN bracis-33639 233 6 of of ADP bracis-33639 233 7 high high ADJ bracis-33639 233 8 and and CCONJ bracis-33639 233 9 low low ADJ bracis-33639 233 10 confidence confidence NOUN bracis-33639 233 11 . . PUNCT bracis-33639 234 1 when when SCONJ bracis-33639 234 2 comparing compare VERB bracis-33639 234 3 traditional traditional ADJ bracis-33639 234 4 machine machine NOUN bracis-33639 234 5 learning learning NOUN bracis-33639 234 6 models model NOUN bracis-33639 234 7 to to PART bracis-33639 234 8 graph graph VERB bracis-33639 234 9 neural neural ADJ bracis-33639 234 10 networks network NOUN bracis-33639 234 11 ( ( PUNCT bracis-33639 234 12 gnns gnns NOUN bracis-33639 234 13 ) ) PUNCT bracis-33639 234 14 , , PUNCT bracis-33639 234 15 the the DET bracis-33639 234 16 latter latter ADJ bracis-33639 234 17 demonstrates demonstrate VERB bracis-33639 234 18 significant significant ADJ bracis-33639 234 19 advantages advantage NOUN bracis-33639 234 20 in in ADP bracis-33639 234 21 capturing capture VERB bracis-33639 234 22 the the DET bracis-33639 234 23 complex complex ADJ bracis-33639 234 24 spatial spatial ADJ bracis-33639 234 25 and and CCONJ bracis-33639 234 26 temporal temporal ADJ bracis-33639 234 27 dependencies dependency NOUN bracis-33639 234 28 inherent inherent ADJ bracis-33639 234 29 in in ADP bracis-33639 234 30 crime crime NOUN bracis-33639 234 31 data datum NOUN bracis-33639 234 32 . . PUNCT bracis-33639 235 1 among among ADP bracis-33639 235 2 the the DET bracis-33639 235 3 traditional traditional ADJ bracis-33639 235 4 models model NOUN bracis-33639 235 5 tested test VERB bracis-33639 235 6 , , PUNCT bracis-33639 235 7 logistic logistic ADJ bracis-33639 235 8 regression regression NOUN bracis-33639 235 9 , , PUNCT bracis-33639 235 10 random random ADJ bracis-33639 235 11 forest forest NOUN bracis-33639 235 12 , , PUNCT bracis-33639 235 13 and and CCONJ bracis-33639 235 14 svm svm ADJ bracis-33639 235 15 , , PUNCT bracis-33639 235 16 random random ADJ bracis-33639 235 17 forest forest NOUN bracis-33639 235 18 achieved achieve VERB bracis-33639 235 19 the the DET bracis-33639 235 20 highest high ADJ bracis-33639 235 21 f1 f1 NOUN bracis-33639 235 22 - - PUNCT bracis-33639 235 23 score score NOUN bracis-33639 235 24 , , PUNCT bracis-33639 235 25 slightly slightly ADV bracis-33639 235 26 surpassing surpass VERB bracis-33639 235 27 the the DET bracis-33639 235 28 gnn gnn NOUN bracis-33639 235 29 models model NOUN bracis-33639 235 30 , , PUNCT bracis-33639 235 31 particularly particularly ADV bracis-33639 235 32 dysat dysat ADJ bracis-33639 235 33 . . PUNCT bracis-33639 236 1 however however ADV bracis-33639 236 2 , , PUNCT bracis-33639 236 3 gnns gnns ADJ bracis-33639 236 4 practical practical ADJ bracis-33639 236 5 application application NOUN bracis-33639 236 6 reveals reveal VERB bracis-33639 236 7 its its PRON bracis-33639 236 8 superiority superiority NOUN bracis-33639 236 9 . . PUNCT bracis-33639 237 1 when when SCONJ bracis-33639 237 2 visualizing visualize VERB bracis-33639 237 3 predictions prediction NOUN bracis-33639 237 4 on on ADP bracis-33639 237 5 a a DET bracis-33639 237 6 street street NOUN bracis-33639 237 7 map map NOUN bracis-33639 237 8 , , PUNCT bracis-33639 237 9 dysat dysat VERB bracis-33639 237 10 effectively effectively ADV bracis-33639 237 11 identifies identify VERB bracis-33639 237 12 crime crime NOUN bracis-33639 237 13 hotspots hotspot NOUN bracis-33639 237 14 with with ADP bracis-33639 237 15 clear clear ADJ bracis-33639 237 16 regions region NOUN bracis-33639 237 17 of of ADP bracis-33639 237 18 red red ADJ bracis-33639 237 19 ( ( PUNCT bracis-33639 237 20 crime crime NOUN bracis-33639 237 21 nodes node NOUN bracis-33639 237 22 ) ) PUNCT bracis-33639 237 23 and and CCONJ bracis-33639 237 24 green green ADJ bracis-33639 237 25 ( ( PUNCT bracis-33639 237 26 no no DET bracis-33639 237 27 crime crime NOUN bracis-33639 237 28 nodes nod VERB bracis-33639 237 29 ) ) PUNCT bracis-33639 237 30 and and CCONJ bracis-33639 237 31 minimal minimal ADJ bracis-33639 237 32 uncertainty uncertainty NOUN bracis-33639 237 33 ( ( PUNCT bracis-33639 237 34 blue blue ADJ bracis-33639 237 35 lines line NOUN bracis-33639 237 36 ) ) PUNCT bracis-33639 237 37 as as SCONJ bracis-33639 237 38 shown show VERB bracis-33639 237 39 in in ADP bracis-33639 237 40 fig fig NOUN bracis-33639 237 41 . . PUNCT bracis-33639 237 42     SPACE bracis-33639 238 1 5 5 NUM bracis-33639 238 2 . . PUNCT bracis-33639 239 1 this this PRON bracis-33639 239 2 contrasts contrast VERB bracis-33639 239 3 with with ADP bracis-33639 239 4 random random ADJ bracis-33639 239 5 forest forest NOUN bracis-33639 239 6 , , PUNCT bracis-33639 239 7 which which PRON bracis-33639 239 8 shows show VERB bracis-33639 239 9 more more ADJ bracis-33639 239 10 uncertainty uncertainty NOUN bracis-33639 239 11 and and CCONJ bracis-33639 239 12 less less ADV bracis-33639 239 13 distinct distinct ADJ bracis-33639 239 14 regions region NOUN bracis-33639 239 15 . . PUNCT bracis-33639 240 1 thus thus ADV bracis-33639 240 2 , , PUNCT bracis-33639 240 3 despite despite SCONJ bracis-33639 240 4 a a DET bracis-33639 240 5 some some DET bracis-33639 240 6 slightly slightly ADV bracis-33639 240 7 lower low ADJ bracis-33639 240 8 scores score NOUN bracis-33639 240 9 , , PUNCT bracis-33639 240 10 gnns gnns ADJ bracis-33639 240 11 practical practical ADJ bracis-33639 240 12 performance performance NOUN bracis-33639 240 13 in in ADP bracis-33639 240 14 hotspot hotspot NOUN bracis-33639 240 15 identification identification NOUN bracis-33639 240 16 demonstrates demonstrate VERB bracis-33639 240 17 its its PRON bracis-33639 240 18 superior superior ADJ bracis-33639 240 19 effectiveness effectiveness NOUN bracis-33639 240 20 over over ADP bracis-33639 240 21 traditional traditional ADJ bracis-33639 240 22 models model NOUN bracis-33639 240 23 . . PUNCT bracis-33639 241 1 while while SCONJ bracis-33639 241 2 somewhat somewhat ADV bracis-33639 241 3 subjective subjective ADJ bracis-33639 241 4 , , PUNCT bracis-33639 241 5 this this DET bracis-33639 241 6 approach approach NOUN bracis-33639 241 7 enhances enhance VERB bracis-33639 241 8 our our PRON bracis-33639 241 9 understanding understanding NOUN bracis-33639 241 10 of of ADP bracis-33639 241 11 spatial spatial ADJ bracis-33639 241 12 crime crime NOUN bracis-33639 241 13 patterns pattern NOUN bracis-33639 241 14 . . PUNCT bracis-33639 242 1 7 7 NUM bracis-33639 242 2 conclusion conclusion NOUN bracis-33639 242 3 this this DET bracis-33639 242 4 study study NOUN bracis-33639 242 5 demonstrates demonstrate VERB bracis-33639 242 6 the the DET bracis-33639 242 7 potential potential NOUN bracis-33639 242 8 of of ADP bracis-33639 242 9 graph graph NOUN bracis-33639 242 10 neural neural ADJ bracis-33639 242 11 networks network NOUN bracis-33639 242 12 ( ( PUNCT bracis-33639 242 13 gnns gnns NOUN bracis-33639 242 14 ) ) PUNCT bracis-33639 242 15 , , PUNCT bracis-33639 242 16 specifically specifically ADV bracis-33639 242 17 dysat dysat VERB bracis-33639 242 18 and and CCONJ bracis-33639 242 19 evolvegcn evolvegcn ADJ bracis-33639 242 20 , , PUNCT bracis-33639 242 21 in in ADP bracis-33639 242 22 improving improve VERB bracis-33639 242 23 crime crime NOUN bracis-33639 242 24 prediction prediction NOUN bracis-33639 242 25 accuracy accuracy NOUN bracis-33639 242 26 by by ADP bracis-33639 242 27 effectively effectively ADV bracis-33639 242 28 modeling model VERB bracis-33639 242 29 crime crime NOUN bracis-33639 242 30 data datum NOUN bracis-33639 242 31 ’s ’s PART bracis-33639 242 32 dynamic dynamic ADJ bracis-33639 242 33 and and CCONJ bracis-33639 242 34 interconnected interconnected ADJ bracis-33639 242 35 nature nature NOUN bracis-33639 242 36 . . PUNCT bracis-33639 243 1 the the DET bracis-33639 243 2 developed develop VERB bracis-33639 243 3 data datum NOUN bracis-33639 243 4 modeling model VERB bracis-33639 243 5 pipeline pipeline NOUN bracis-33639 243 6 , , PUNCT bracis-33639 243 7 integrating integrate VERB bracis-33639 243 8 crime crime NOUN bracis-33639 243 9 data datum NOUN bracis-33639 243 10 with with ADP bracis-33639 243 11 street street NOUN bracis-33639 243 12 map map NOUN bracis-33639 243 13 graphs graph NOUN bracis-33639 243 14 , , PUNCT bracis-33639 243 15 enables enable VERB bracis-33639 243 16 a a DET bracis-33639 243 17 detailed detailed ADJ bracis-33639 243 18 representation representation NOUN bracis-33639 243 19 of of ADP bracis-33639 243 20 crime crime NOUN bracis-33639 243 21 dynamics dynamic NOUN bracis-33639 243 22 in in ADP bracis-33639 243 23 são são PROPN bracis-33639 243 24 paulo paulo PROPN bracis-33639 243 25 . . PUNCT bracis-33639 244 1 our our PRON bracis-33639 244 2 evaluation evaluation NOUN bracis-33639 244 3 and and CCONJ bracis-33639 244 4 comparison comparison NOUN bracis-33639 244 5 of of ADP bracis-33639 244 6 dysat dysat NOUN bracis-33639 244 7 and and CCONJ bracis-33639 244 8 evolvegcn evolvegcn PROPN bracis-33639 244 9 reveal reveal VERB bracis-33639 244 10 their their PRON bracis-33639 244 11 effectiveness effectiveness NOUN bracis-33639 244 12 in in ADP bracis-33639 244 13 learning learn VERB bracis-33639 244 14 from from ADP bracis-33639 244 15 historical historical ADJ bracis-33639 244 16 data datum NOUN bracis-33639 244 17 and and CCONJ bracis-33639 244 18 predicting predict VERB bracis-33639 244 19 future future ADJ bracis-33639 244 20 crime crime NOUN bracis-33639 244 21 occurrences occurrence NOUN bracis-33639 244 22 . . PUNCT bracis-33639 245 1 these these DET bracis-33639 245 2 findings finding NOUN bracis-33639 245 3 highlight highlight VERB bracis-33639 245 4 the the DET bracis-33639 245 5 need need NOUN bracis-33639 245 6 for for ADP bracis-33639 245 7 further further ADJ bracis-33639 245 8 exploration exploration NOUN bracis-33639 245 9 and and CCONJ bracis-33639 245 10 adoption adoption NOUN bracis-33639 245 11 of of ADP bracis-33639 245 12 gnns gnns NOUN bracis-33639 245 13 in in ADP bracis-33639 245 14 the the DET bracis-33639 245 15 crime crime NOUN bracis-33639 245 16 prediction prediction NOUN bracis-33639 245 17 domain domain NOUN bracis-33639 245 18 , , PUNCT bracis-33639 245 19 offering offer VERB bracis-33639 245 20 valuable valuable ADJ bracis-33639 245 21 insights insight NOUN bracis-33639 245 22 for for ADP bracis-33639 245 23 enhancing enhance VERB bracis-33639 245 24 public public ADJ bracis-33639 245 25 safety safety NOUN bracis-33639 245 26 strategies strategy NOUN bracis-33639 245 27 and and CCONJ bracis-33639 245 28 crime crime NOUN bracis-33639 245 29 prevention prevention NOUN bracis-33639 245 30 efforts effort NOUN bracis-33639 245 31 . . PUNCT bracis-33639 246 1 notes note NOUN bracis-33639 246 2 1.https://github.com/giva-lab/bracis2024_gnn.git 1.https://github.com/giva-lab/bracis2024_gnn.git PROPN bracis-33639 246 3 . . PUNCT bracis-33639 247 1 references reference NOUN bracis-33639 247 2 ahishakiye ahishakiye PROPN bracis-33639 247 3 , , PUNCT bracis-33639 247 4 e. e. PROPN bracis-33639 247 5 , , PUNCT bracis-33639 247 6 taremwa taremwa PROPN bracis-33639 247 7 , , PUNCT bracis-33639 247 8 d. d. PROPN bracis-33639 247 9 , , PUNCT bracis-33639 247 10 omulo omulo PROPN bracis-33639 247 11 , , PUNCT bracis-33639 247 12 e.o e.o PROPN bracis-33639 247 13 . . PROPN bracis-33639 247 14 , , PUNCT bracis-33639 247 15 niyonzima niyonzima PROPN bracis-33639 247 16 , , PUNCT bracis-33639 247 17 i. i. NOUN bracis-33639 247 18 : : PUNCT bracis-33639 247 19 crime crime NOUN bracis-33639 247 20 prediction prediction NOUN bracis-33639 247 21 using use VERB bracis-33639 247 22 decision decision NOUN bracis-33639 247 23 tree tree NOUN bracis-33639 247 24 ( ( PUNCT bracis-33639 247 25 j48 j48 ADJ bracis-33639 247 26 ) ) PUNCT bracis-33639 247 27 classification classification NOUN bracis-33639 247 28 algorithm algorithm NOUN bracis-33639 247 29 . . PUNCT bracis-33639 248 1 int int NOUN bracis-33639 248 2 . . PUNCT bracis-33639 249 1 j. j. PROPN bracis-33639 249 2 comput comput PROPN bracis-33639 249 3 . . PUNCT bracis-33639 250 1 inf inf PROPN bracis-33639 250 2 . . PUNCT bracis-33639 250 3 technol technol PROPN bracis-33639 250 4 . . PUNCT bracis-33639 250 5 6(3 6(3 NUM bracis-33639 250 6 ) ) PUNCT bracis-33639 250 7 , , PUNCT bracis-33639 250 8 188–195 188–195 NUM bracis-33639 250 9 ( ( PUNCT bracis-33639 250 10 2017 2017 NUM bracis-33639 250 11 ) ) PUNCT bracis-33639 250 12 google google PROPN bracis-33639 250 13 scholar scholar NOUN bracis-33639 250 14     SPACE bracis-33639 250 15 ahmed ahmed PROPN bracis-33639 250 16 , , PUNCT bracis-33639 250 17 s. s. PROPN bracis-33639 250 18 , , PUNCT bracis-33639 250 19 gentili gentili PROPN bracis-33639 250 20 , , PUNCT bracis-33639 250 21 m. m. NOUN bracis-33639 250 22 , , PUNCT bracis-33639 250 23 sierra sierra PROPN bracis-33639 250 24 - - PUNCT bracis-33639 250 25 sosa sosa PROPN bracis-33639 250 26 , , PUNCT bracis-33639 250 27 d. d. PROPN bracis-33639 250 28 , , PUNCT bracis-33639 250 29 elmaghraby elmaghraby PROPN bracis-33639 250 30 , , PUNCT bracis-33639 250 31 a.s a.s PROPN bracis-33639 250 32 . . PROPN bracis-33639 250 33 : : PUNCT bracis-33639 251 1 multi multi ADJ bracis-33639 251 2 - - ADJ bracis-33639 251 3 layer layer ADJ bracis-33639 251 4 data datum NOUN bracis-33639 251 5 integration integration NOUN bracis-33639 251 6 technique technique NOUN bracis-33639 251 7 for for ADP bracis-33639 251 8 combining combine VERB bracis-33639 251 9 heterogeneous heterogeneous ADJ bracis-33639 251 10 crime crime NOUN bracis-33639 251 11 data datum NOUN bracis-33639 251 12 . . PUNCT bracis-33639 252 1 inf inf PROPN bracis-33639 252 2 . . PROPN bracis-33639 252 3 process process NOUN bracis-33639 252 4 . . PUNCT bracis-33639 253 1 manag manag PROPN bracis-33639 253 2 . . PUNCT bracis-33639 254 1 59(3 59(3 NUM bracis-33639 254 2 ) ) PUNCT bracis-33639 254 3 , , PUNCT bracis-33639 254 4 102879 102879 NUM bracis-33639 254 5 ( ( PUNCT bracis-33639 254 6 2022 2022 NUM bracis-33639 254 7 ) ) PUNCT bracis-33639 254 8 article article NOUN bracis-33639 254 9     SPACE bracis-33639 254 10 math math NOUN bracis-33639 254 11     SPACE bracis-33639 254 12 google google PROPN bracis-33639 254 13 scholar scholar NOUN bracis-33639 254 14     SPACE bracis-33639 254 15 boeing boeing PROPN bracis-33639 254 16 , , PUNCT bracis-33639 254 17 g. g. PROPN bracis-33639 254 18 : : PUNCT bracis-33639 254 19 modeling modeling NOUN bracis-33639 254 20 and and CCONJ bracis-33639 254 21 analyzing analyze VERB bracis-33639 254 22 urban urban ADJ bracis-33639 254 23 networks network NOUN bracis-33639 254 24 and and CCONJ bracis-33639 254 25 amenities amenity NOUN bracis-33639 254 26 with with ADP bracis-33639 254 27 osmnx osmnx PROPN bracis-33639 254 28 ( ( PUNCT bracis-33639 254 29 2024 2024 NUM bracis-33639 254 30 ) ) PUNCT bracis-33639 254 31 . . PUNCT bracis-33639 255 1 https://geoffboeing.com/share/osmnx-paper.pdf https://geoffboeing.com/share/osmnx-paper.pdf PROPN bracis-33639 255 2 casali casali PROPN bracis-33639 255 3 , , PUNCT bracis-33639 255 4 y. y. NOUN bracis-33639 255 5 , , PUNCT bracis-33639 255 6 aydin aydin PROPN bracis-33639 255 7 , , PUNCT bracis-33639 255 8 n.y n.y PROPN bracis-33639 255 9 . . PROPN bracis-33639 255 10 , , PUNCT bracis-33639 255 11 comes come VERB bracis-33639 255 12 , , PUNCT bracis-33639 255 13 t. t. PROPN bracis-33639 255 14 : : PUNCT bracis-33639 255 15 machine machine NOUN bracis-33639 255 16 learning learning NOUN bracis-33639 255 17 for for ADP bracis-33639 255 18 spatial spatial ADJ bracis-33639 255 19 analyses analysis NOUN bracis-33639 255 20 in in ADP bracis-33639 255 21 urban urban ADJ bracis-33639 255 22 areas area NOUN bracis-33639 255 23 : : PUNCT bracis-33639 255 24 a a DET bracis-33639 255 25 scoping scope VERB bracis-33639 255 26 review review NOUN bracis-33639 255 27 . . PUNCT bracis-33639 256 1 sustain sustain VERB bracis-33639 256 2 . . PUNCT bracis-33639 257 1 urban urban ADJ bracis-33639 257 2 areas area NOUN bracis-33639 257 3 85 85 NUM bracis-33639 257 4 , , PUNCT bracis-33639 257 5 104050 104050 NUM bracis-33639 257 6 ( ( PUNCT bracis-33639 257 7 2022 2022 NUM bracis-33639 257 8 ) ) PUNCT bracis-33639 257 9 math math NOUN bracis-33639 257 10     SPACE bracis-33639 257 11 google google PROPN bracis-33639 257 12 scholar scholar NOUN bracis-33639 257 13     SPACE bracis-33639 257 14 dakalbab dakalbab PROPN bracis-33639 257 15 , , PUNCT bracis-33639 257 16 f. f. PROPN bracis-33639 257 17 , , PUNCT bracis-33639 257 18 talib talib PROPN bracis-33639 257 19 , , PUNCT bracis-33639 257 20 m.a m.a PROPN bracis-33639 257 21 . . PROPN bracis-33639 257 22 , , PUNCT bracis-33639 257 23 waraga waraga PROPN bracis-33639 257 24 , , PUNCT bracis-33639 257 25 o.a o.a PROPN bracis-33639 257 26 . . PROPN bracis-33639 257 27 , , PUNCT bracis-33639 257 28 nassif nassif NOUN bracis-33639 257 29 , , PUNCT bracis-33639 257 30 a.b a.b PROPN bracis-33639 257 31 . . PROPN bracis-33639 257 32 , , PUNCT bracis-33639 257 33 abbas abbas PROPN bracis-33639 257 34 , , PUNCT bracis-33639 257 35 s. s. PROPN bracis-33639 257 36 , , PUNCT bracis-33639 257 37 nasir nasir PROPN bracis-33639 257 38 , , PUNCT bracis-33639 257 39 q. q. PROPN bracis-33639 257 40 : : PUNCT bracis-33639 257 41 artificial artificial ADJ bracis-33639 257 42 intelligence intelligence PROPN bracis-33639 257 43 & & CCONJ bracis-33639 257 44 crime crime PROPN bracis-33639 257 45 prediction prediction PROPN bracis-33639 257 46 : : PUNCT bracis-33639 257 47 a a DET bracis-33639 257 48 systematic systematic ADJ bracis-33639 257 49 literature literature PROPN bracis-33639 257 50 review review NOUN bracis-33639 257 51 . . PUNCT bracis-33639 258 1 soc soc PROPN bracis-33639 258 2 . . PUNCT bracis-33639 259 1 sci sci PROPN bracis-33639 259 2 . . PUNCT bracis-33639 259 3 humanit humanit PROPN bracis-33639 259 4 . . PUNCT bracis-33639 260 1 open open VERB bracis-33639 260 2 6(1 6(1 NUM bracis-33639 260 3 ) ) PUNCT bracis-33639 260 4 , , PUNCT bracis-33639 260 5 100342 100342 NUM bracis-33639 260 6 ( ( PUNCT bracis-33639 260 7 2022 2022 NUM bracis-33639 260 8 ) ) PUNCT bracis-33639 260 9 google google PROPN bracis-33639 260 10 scholar scholar NOUN bracis-33639 260 11     SPACE bracis-33639 260 12 eck eck PROPN bracis-33639 260 13 , , PUNCT bracis-33639 260 14 j. j. PROPN bracis-33639 260 15 , , PUNCT bracis-33639 260 16 chainey chainey NOUN bracis-33639 260 17 , , PUNCT bracis-33639 260 18 s. s. PROPN bracis-33639 260 19 , , PUNCT bracis-33639 260 20 cameron cameron PROPN bracis-33639 260 21 , , PUNCT bracis-33639 260 22 j. j. PROPN bracis-33639 260 23 , , PUNCT bracis-33639 260 24 wilson wilson PROPN bracis-33639 260 25 , , PUNCT bracis-33639 260 26 r. r. PROPN bracis-33639 260 27 : : PUNCT bracis-33639 261 1 mapping map VERB bracis-33639 261 2 crime crime NOUN bracis-33639 261 3 : : PUNCT bracis-33639 261 4 understanding understand VERB bracis-33639 261 5 hotspots hotspot NOUN bracis-33639 261 6 ( ( PUNCT bracis-33639 261 7 2005 2005 NUM bracis-33639 261 8 ) ) PUNCT bracis-33639 261 9 google google PROPN bracis-33639 261 10 scholar scholar NOUN bracis-33639 261 11     SPACE bracis-33639 261 12 fan fan PROPN bracis-33639 261 13 , , PUNCT bracis-33639 261 14 w. w. PROPN bracis-33639 261 15 , , PUNCT bracis-33639 261 16 et et PROPN bracis-33639 261 17 al al PROPN bracis-33639 261 18 . . PUNCT bracis-33639 261 19 : : PUNCT bracis-33639 261 20 graph graph VERB bracis-33639 261 21 neural neural ADJ bracis-33639 261 22 networks network NOUN bracis-33639 261 23 for for ADP bracis-33639 261 24 social social ADJ bracis-33639 261 25 recommendation recommendation NOUN bracis-33639 261 26 . . PUNCT bracis-33639 262 1 in in ADP bracis-33639 262 2 : : PUNCT bracis-33639 262 3 the the DET bracis-33639 262 4 world world NOUN bracis-33639 262 5 wide wide ADJ bracis-33639 262 6 web web NOUN bracis-33639 262 7 conference conference NOUN bracis-33639 262 8 , , PUNCT bracis-33639 262 9 www www NOUN bracis-33639 262 10 2019 2019 NUM bracis-33639 262 11 , , PUNCT bracis-33639 262 12 pp pp ADJ bracis-33639 262 13 . . PUNCT bracis-33639 263 1 417–426 417–426 NUM bracis-33639 263 2 . . PUNCT bracis-33639 263 3 association association NOUN bracis-33639 263 4 for for ADP bracis-33639 263 5 computing compute VERB bracis-33639 263 6 machinery machinery NOUN bracis-33639 263 7 , , PUNCT bracis-33639 263 8 new new PROPN bracis-33639 263 9 york york PROPN bracis-33639 263 10 ( ( PUNCT bracis-33639 263 11 2019 2019 NUM bracis-33639 263 12 ) ) PUNCT bracis-33639 263 13 . . PUNCT bracis-33639 264 1 https://doi.org/10.1145/3308558.3313488 https://doi.org/10.1145/3308558.3313488 X bracis-33639 264 2 garcía garcía NOUN bracis-33639 264 3 - - PUNCT bracis-33639 264 4 zanabria zanabria PROPN bracis-33639 264 5 , , PUNCT bracis-33639 264 6 g. g. PROPN bracis-33639 264 7 , , PUNCT bracis-33639 264 8 et et PROPN bracis-33639 264 9 al al PROPN bracis-33639 264 10 . . PUNCT bracis-33639 264 11 : : PUNCT bracis-33639 265 1 cripav cripav NOUN bracis-33639 265 2 : : PUNCT bracis-33639 265 3 street street NOUN bracis-33639 265 4 - - PUNCT bracis-33639 265 5 level level NOUN bracis-33639 265 6 crime crime NOUN bracis-33639 265 7 patterns pattern VERB bracis-33639 265 8 analysis analysis NOUN bracis-33639 265 9 and and CCONJ bracis-33639 265 10 visualization visualization NOUN bracis-33639 265 11 . . PUNCT bracis-33639 266 1 ieee ieee PROPN bracis-33639 266 2 trans trans PROPN bracis-33639 266 3 . . PUNCT bracis-33639 267 1 vis vis X bracis-33639 267 2 . . PUNCT bracis-33639 268 1 comput comput NOUN bracis-33639 268 2 . . PUNCT bracis-33639 269 1 graph graph NOUN bracis-33639 269 2 . . PUNCT bracis-33639 270 1 28(12 28(12 NUM bracis-33639 270 2 ) ) PUNCT bracis-33639 270 3 , , PUNCT bracis-33639 270 4 4000–4015 4000–4015 NUM bracis-33639 270 5 ( ( PUNCT bracis-33639 270 6 2022 2022 NUM bracis-33639 270 7 ) ) PUNCT bracis-33639 270 8 . . PUNCT bracis-33639 271 1 https://doi.org/10.1109/tvcg.2021.3111146 https://doi.org/10.1109/tvcg.2021.3111146 PROPN bracis-33639 271 2 gorr gorr NOUN bracis-33639 271 3 , , PUNCT bracis-33639 271 4 w. w. PROPN bracis-33639 271 5 , , PUNCT bracis-33639 271 6 harries harries PROPN bracis-33639 271 7 , , PUNCT bracis-33639 271 8 r. r. PROPN bracis-33639 271 9 : : PUNCT bracis-33639 271 10 introduction introduction NOUN bracis-33639 271 11 to to ADP bracis-33639 271 12 crime crime NOUN bracis-33639 271 13 forecasting forecasting NOUN bracis-33639 271 14 . . PUNCT bracis-33639 272 1 int int NOUN bracis-33639 272 2 . . PUNCT bracis-33639 273 1 j. j. PROPN bracis-33639 273 2 forecast forecast PROPN bracis-33639 273 3 . . PUNCT bracis-33639 274 1 19(4 19(4 PROPN bracis-33639 274 2 ) ) PUNCT bracis-33639 274 3 , , PUNCT bracis-33639 275 1 551–555 551–555 NUM bracis-33639 275 2 ( ( PUNCT bracis-33639 275 3 2003 2003 NUM bracis-33639 275 4 ) ) PUNCT bracis-33639 275 5 . . PUNCT bracis-33639 276 1 https://doi.org/10.1016/s0169-2070(03)00089-x https://doi.org/10.1016/s0169-2070(03)00089-x PROPN bracis-33639 276 2 article article NOUN bracis-33639 276 3     SPACE bracis-33639 276 4 math math NOUN bracis-33639 276 5     SPACE bracis-33639 276 6 google google PROPN bracis-33639 276 7 scholar scholar NOUN bracis-33639 276 8     SPACE bracis-33639 276 9 han han PROPN bracis-33639 276 10 , , PUNCT bracis-33639 276 11 x. x. PROPN bracis-33639 276 12 , , PUNCT bracis-33639 276 13 hu hu PROPN bracis-33639 276 14 , , PUNCT bracis-33639 276 15 x. x. PROPN bracis-33639 276 16 , , PUNCT bracis-33639 276 17 wu wu PROPN bracis-33639 276 18 , , PUNCT bracis-33639 276 19 h. h. PROPN bracis-33639 276 20 , , PUNCT bracis-33639 276 21 shen shen PROPN bracis-33639 276 22 , , PUNCT bracis-33639 276 23 b. b. PROPN bracis-33639 276 24 , , PUNCT bracis-33639 276 25 wu wu PROPN bracis-33639 276 26 , , PUNCT bracis-33639 276 27 j. j. PROPN bracis-33639 276 28 : : PUNCT bracis-33639 276 29 risk risk NOUN bracis-33639 276 30 prediction prediction NOUN bracis-33639 276 31 of of ADP bracis-33639 276 32 theft theft NOUN bracis-33639 276 33 crimes crime NOUN bracis-33639 276 34 in in ADP bracis-33639 276 35 urban urban ADJ bracis-33639 276 36 communities community NOUN bracis-33639 276 37 : : PUNCT bracis-33639 276 38 an an DET bracis-33639 276 39 integrated integrated ADJ bracis-33639 276 40 model model NOUN bracis-33639 276 41 of of ADP bracis-33639 276 42 lstm lstm PROPN bracis-33639 276 43 and and CCONJ bracis-33639 276 44 st st PROPN bracis-33639 276 45 - - PUNCT bracis-33639 276 46 gcn gcn NOUN bracis-33639 276 47 . . PUNCT bracis-33639 277 1 ieee ieee NOUN bracis-33639 277 2 access access NOUN bracis-33639 277 3 8 8 NUM bracis-33639 277 4 , , PUNCT bracis-33639 277 5 217222–217230 217222–217230 NUM bracis-33639 277 6 ( ( PUNCT bracis-33639 277 7 2020 2020 NUM bracis-33639 277 8 ) ) PUNCT bracis-33639 277 9 . . PUNCT bracis-33639 278 1 https://doi.org/10.1109/access.2020.3041924 https://doi.org/10.1109/access.2020.3041924 PROPN bracis-33639 278 2 article article PROPN bracis-33639 278 3     SPACE bracis-33639 278 4 google google PROPN bracis-33639 278 5 scholar scholar NOUN bracis-33639 278 6     SPACE bracis-33639 278 7 jenga jenga NOUN bracis-33639 278 8 , , PUNCT bracis-33639 278 9 k. k. PROPN bracis-33639 278 10 , , PUNCT bracis-33639 278 11 catal catal PROPN bracis-33639 278 12 , , PUNCT bracis-33639 278 13 c. c. PROPN bracis-33639 278 14 , , PUNCT bracis-33639 278 15 kar kar PROPN bracis-33639 278 16 , , PUNCT bracis-33639 278 17 g. g. NOUN bracis-33639 278 18 : : PUNCT bracis-33639 278 19 machine machine NOUN bracis-33639 278 20 learning learning NOUN bracis-33639 278 21 in in ADP bracis-33639 278 22 crime crime NOUN bracis-33639 278 23 prediction prediction NOUN bracis-33639 278 24 . . PUNCT bracis-33639 279 1 j. j. PROPN bracis-33639 279 2 ambient ambient PROPN bracis-33639 279 3 . . PUNCT bracis-33639 280 1 intell intell PROPN bracis-33639 280 2 . . PUNCT bracis-33639 281 1 humaniz humaniz VERB bracis-33639 281 2 . . PUNCT bracis-33639 282 1 comput comput NOUN bracis-33639 282 2 . . PUNCT bracis-33639 283 1 14(3 14(3 NUM bracis-33639 283 2 ) ) PUNCT bracis-33639 283 3 , , PUNCT bracis-33639 283 4 2887–2913 2887–2913 NUM bracis-33639 283 5 ( ( PUNCT bracis-33639 283 6 2023 2023 NUM bracis-33639 283 7 ) ) PUNCT bracis-33639 283 8 article article NOUN bracis-33639 283 9     SPACE bracis-33639 283 10 math math NOUN bracis-33639 283 11     SPACE bracis-33639 283 12 google google PROPN bracis-33639 283 13 scholar scholar NOUN bracis-33639 283 14     SPACE bracis-33639 283 15 jiang jiang PROPN bracis-33639 283 16 , , PUNCT bracis-33639 283 17 w. w. PROPN bracis-33639 283 18 , , PUNCT bracis-33639 283 19 luo luo PROPN bracis-33639 283 20 , , PUNCT bracis-33639 283 21 j. j. PROPN bracis-33639 283 22 : : PUNCT bracis-33639 283 23 graph graph VERB bracis-33639 283 24 neural neural ADJ bracis-33639 283 25 network network NOUN bracis-33639 283 26 for for ADP bracis-33639 283 27 traffic traffic NOUN bracis-33639 283 28 forecasting forecasting NOUN bracis-33639 283 29 : : PUNCT bracis-33639 283 30 a a DET bracis-33639 283 31 survey survey NOUN bracis-33639 283 32 . . PUNCT bracis-33639 284 1 expert expert NOUN bracis-33639 284 2 syst syst PROPN bracis-33639 284 3 . . PUNCT bracis-33639 285 1 appl appl PROPN bracis-33639 285 2 . . PROPN bracis-33639 285 3 207 207 NUM bracis-33639 285 4 , , PUNCT bracis-33639 285 5 117921 117921 NUM bracis-33639 285 6 ( ( PUNCT bracis-33639 285 7 2022 2022 NUM bracis-33639 285 8 ) ) PUNCT bracis-33639 285 9 . . PUNCT bracis-33639 286 1 https://www.sciencedirect.com/science/article/pii/s0957417422011654 https://www.sciencedirect.com/science/article/pii/s0957417422011654 PROPN bracis-33639 287 1 jin jin PROPN bracis-33639 287 2 , , PUNCT bracis-33639 287 3 g. g. PROPN bracis-33639 287 4 , , PUNCT bracis-33639 287 5 wang wang PROPN bracis-33639 287 6 , , PUNCT bracis-33639 287 7 q. q. PROPN bracis-33639 287 8 , , PUNCT bracis-33639 287 9 zhu zhu PROPN bracis-33639 287 10 , , PUNCT bracis-33639 287 11 c. c. PROPN bracis-33639 287 12 , , PUNCT bracis-33639 287 13 feng feng PROPN bracis-33639 287 14 , , PUNCT bracis-33639 287 15 y. y. PROPN bracis-33639 287 16 , , PUNCT bracis-33639 287 17 huang huang PROPN bracis-33639 287 18 , , PUNCT bracis-33639 287 19 j. j. PROPN bracis-33639 287 20 , , PUNCT bracis-33639 287 21 zhou zhou PROPN bracis-33639 287 22 , , PUNCT bracis-33639 287 23 j. j. PROPN bracis-33639 287 24 : : PUNCT bracis-33639 287 25 addressing address VERB bracis-33639 287 26 crime crime NOUN bracis-33639 287 27 situation situation NOUN bracis-33639 287 28 forecasting forecasting NOUN bracis-33639 287 29 task task NOUN bracis-33639 287 30 with with ADP bracis-33639 287 31 temporal temporal ADJ bracis-33639 287 32 graph graph NOUN bracis-33639 287 33 convolutional convolutional ADJ bracis-33639 287 34 neural neural ADJ bracis-33639 287 35 network network NOUN bracis-33639 287 36 approach approach NOUN bracis-33639 287 37 . . PUNCT bracis-33639 288 1 in in ADP bracis-33639 288 2 : : PUNCT bracis-33639 288 3 2020 2020 NUM bracis-33639 288 4 12th 12th ADJ bracis-33639 288 5 international international ADJ bracis-33639 288 6 conference conference NOUN bracis-33639 288 7 on on ADP bracis-33639 288 8 measuring measure VERB bracis-33639 288 9 technology technology NOUN bracis-33639 288 10 and and CCONJ bracis-33639 288 11 mechatronics mechatronic NOUN bracis-33639 288 12 automation automation NOUN bracis-33639 288 13 ( ( PUNCT bracis-33639 288 14 icmtma icmtma NOUN bracis-33639 288 15 ) ) PUNCT bracis-33639 288 16 , , PUNCT bracis-33639 288 17 pp pp ADP bracis-33639 288 18 . . PUNCT bracis-33639 289 1 474–478 474–478 NUM bracis-33639 289 2 ( ( PUNCT bracis-33639 289 3 2020 2020 NUM bracis-33639 289 4 ) ) PUNCT bracis-33639 289 5 . . PUNCT bracis-33639 290 1 https://doi.org/10.1109/icmtma50254.2020.00108 https://doi.org/10.1109/icmtma50254.2020.00108 NUM bracis-33639 290 2 jin jin PROPN bracis-33639 290 3 , , PUNCT bracis-33639 290 4 g. g. PROPN bracis-33639 290 5 , , PUNCT bracis-33639 290 6 wang wang PROPN bracis-33639 290 7 , , PUNCT bracis-33639 290 8 q. q. PROPN bracis-33639 290 9 , , PUNCT bracis-33639 290 10 zhu zhu PROPN bracis-33639 290 11 , , PUNCT bracis-33639 290 12 c. c. PROPN bracis-33639 290 13 , , PUNCT bracis-33639 290 14 feng feng PROPN bracis-33639 290 15 , , PUNCT bracis-33639 290 16 y. y. PROPN bracis-33639 290 17 , , PUNCT bracis-33639 290 18 huang huang PROPN bracis-33639 290 19 , , PUNCT bracis-33639 290 20 j. j. PROPN bracis-33639 290 21 , , PUNCT bracis-33639 290 22 zhou zhou PROPN bracis-33639 290 23 , , PUNCT bracis-33639 290 24 j. j. PROPN bracis-33639 290 25 : : PUNCT bracis-33639 290 26 addressing address VERB bracis-33639 290 27 crime crime NOUN bracis-33639 290 28 situation situation NOUN bracis-33639 290 29 forecasting forecasting NOUN bracis-33639 290 30 task task NOUN bracis-33639 290 31 with with ADP bracis-33639 290 32 temporal temporal ADJ bracis-33639 290 33 graph graph NOUN bracis-33639 290 34 convolutional convolutional ADJ bracis-33639 290 35 neural neural ADJ bracis-33639 290 36 network network NOUN bracis-33639 290 37 approach approach NOUN bracis-33639 290 38 . . PUNCT bracis-33639 291 1 in in ADP bracis-33639 291 2 : : PUNCT bracis-33639 291 3 2020 2020 NUM bracis-33639 291 4 12th 12th ADJ bracis-33639 291 5 international international ADJ bracis-33639 291 6 conference conference NOUN bracis-33639 291 7 on on ADP bracis-33639 291 8 measuring measure VERB bracis-33639 291 9 technology technology NOUN bracis-33639 291 10 and and CCONJ bracis-33639 291 11 mechatronics mechatronic NOUN bracis-33639 291 12 automation automation NOUN bracis-33639 291 13 ( ( PUNCT bracis-33639 291 14 icmtma icmtma NOUN bracis-33639 291 15 ) ) PUNCT bracis-33639 291 16 , , PUNCT bracis-33639 291 17 pp pp ADP bracis-33639 291 18 . . PUNCT bracis-33639 292 1 474–478 474–478 NUM bracis-33639 292 2 ( ( PUNCT bracis-33639 292 3 2020 2020 NUM bracis-33639 292 4 ) ) PUNCT bracis-33639 292 5 google google PROPN bracis-33639 292 6 scholar scholar NOUN bracis-33639 292 7     SPACE bracis-33639 292 8 kipf kipf PROPN bracis-33639 292 9 , , PUNCT bracis-33639 292 10 t.n t.n PROPN bracis-33639 292 11 . . PROPN bracis-33639 292 12 , , PUNCT bracis-33639 292 13 welling well VERB bracis-33639 292 14 , , PUNCT bracis-33639 292 15 m. m. NOUN bracis-33639 292 16 : : PUNCT bracis-33639 292 17 semi semi ADJ bracis-33639 292 18 - - ADJ bracis-33639 292 19 supervised supervised ADJ bracis-33639 292 20 classification classification NOUN bracis-33639 292 21 with with ADP bracis-33639 292 22 graph graph NOUN bracis-33639 292 23 convolutional convolutional ADJ bracis-33639 292 24 networks network NOUN bracis-33639 292 25 . . PUNCT bracis-33639 293 1 arxiv arxiv PROPN bracis-33639 293 2 preprint preprint VERB bracis-33639 293 3 arxiv:1609.02907 arxiv:1609.02907 NOUN bracis-33639 293 4 ( ( PUNCT bracis-33639 293 5 2016 2016 NUM bracis-33639 293 6 ) ) PUNCT bracis-33639 293 7 longa longa NOUN bracis-33639 293 8 , , PUNCT bracis-33639 293 9 a. a. PROPN bracis-33639 293 10 , , PUNCT bracis-33639 293 11 et et PROPN bracis-33639 293 12 al al PROPN bracis-33639 293 13 . . PUNCT bracis-33639 293 14 : : PUNCT bracis-33639 293 15 graph graph VERB bracis-33639 293 16 neural neural ADJ bracis-33639 293 17 networks network NOUN bracis-33639 293 18 for for ADP bracis-33639 293 19 temporal temporal ADJ bracis-33639 293 20 graphs graph NOUN bracis-33639 293 21 : : PUNCT bracis-33639 293 22 state state NOUN bracis-33639 293 23 of of ADP bracis-33639 293 24 the the DET bracis-33639 293 25 art art NOUN bracis-33639 293 26 , , PUNCT bracis-33639 293 27 open open ADJ bracis-33639 293 28 challenges challenge NOUN bracis-33639 293 29 , , PUNCT bracis-33639 293 30 and and CCONJ bracis-33639 293 31 opportunities opportunity NOUN bracis-33639 293 32 ( ( PUNCT bracis-33639 293 33 2023 2023 NUM bracis-33639 293 34 ) ) PUNCT bracis-33639 293 35 google google NOUN bracis-33639 293 36 scholar scholar NOUN bracis-33639 293 37     SPACE bracis-33639 293 38 mandalapu mandalapu PROPN bracis-33639 293 39 , , PUNCT bracis-33639 293 40 v. v. ADV bracis-33639 293 41 , , PUNCT bracis-33639 293 42 elluri elluri ADJ bracis-33639 293 43 , , PUNCT bracis-33639 293 44 l. l. PROPN bracis-33639 293 45 , , PUNCT bracis-33639 293 46 vyas vyas PROPN bracis-33639 293 47 , , PUNCT bracis-33639 293 48 p. p. PROPN bracis-33639 293 49 , , PUNCT bracis-33639 293 50 roy roy PROPN bracis-33639 293 51 , , PUNCT bracis-33639 293 52 n. n. NOUN bracis-33639 293 53 : : PUNCT bracis-33639 293 54 crime crime NOUN bracis-33639 293 55 prediction prediction NOUN bracis-33639 293 56 using use VERB bracis-33639 293 57 machine machine NOUN bracis-33639 293 58 learning learning NOUN bracis-33639 293 59 and and CCONJ bracis-33639 293 60 deep deep ADJ bracis-33639 293 61 learning learning NOUN bracis-33639 293 62 : : PUNCT bracis-33639 293 63 a a DET bracis-33639 293 64 systematic systematic ADJ bracis-33639 293 65 review review NOUN bracis-33639 293 66 and and CCONJ bracis-33639 293 67 future future ADJ bracis-33639 293 68 directions direction NOUN bracis-33639 293 69 . . PUNCT bracis-33639 294 1 ieee ieee NOUN bracis-33639 294 2 access access NOUN bracis-33639 294 3 11 11 NUM bracis-33639 294 4 , , PUNCT bracis-33639 294 5 60153–60170 60153–60170 NUM bracis-33639 294 6 ( ( PUNCT bracis-33639 294 7 2023 2023 NUM bracis-33639 294 8 ) ) PUNCT bracis-33639 294 9 . . PUNCT bracis-33639 295 1 https://doi.org/10.1109/access.2023.3286344 https://doi.org/10.1109/access.2023.3286344 PROPN bracis-33639 295 2 article article NOUN bracis-33639 295 3     SPACE bracis-33639 295 4 google google PROPN bracis-33639 295 5 scholar scholar NOUN bracis-33639 295 6     SPACE bracis-33639 295 7 mandalapu mandalapu PROPN bracis-33639 295 8 , , PUNCT bracis-33639 295 9 v. v. ADV bracis-33639 295 10 , , PUNCT bracis-33639 295 11 elluri elluri ADJ bracis-33639 295 12 , , PUNCT bracis-33639 295 13 l. l. PROPN bracis-33639 295 14 , , PUNCT bracis-33639 295 15 vyas vyas PROPN bracis-33639 295 16 , , PUNCT bracis-33639 295 17 p. p. PROPN bracis-33639 295 18 , , PUNCT bracis-33639 295 19 roy roy PROPN bracis-33639 295 20 , , PUNCT bracis-33639 295 21 n. n. NOUN bracis-33639 295 22 : : PUNCT bracis-33639 296 1 crime crime NOUN bracis-33639 296 2 prediction prediction NOUN bracis-33639 296 3 using use VERB bracis-33639 296 4 machine machine NOUN bracis-33639 296 5 learning learning NOUN bracis-33639 296 6 and and CCONJ bracis-33639 296 7 deep deep ADJ bracis-33639 296 8 learning learning NOUN bracis-33639 296 9 : : PUNCT bracis-33639 296 10 a a DET bracis-33639 296 11 systematic systematic ADJ bracis-33639 296 12 review review NOUN bracis-33639 296 13 and and CCONJ bracis-33639 296 14 future future ADJ bracis-33639 296 15 directions direction NOUN bracis-33639 296 16 . . PUNCT bracis-33639 297 1 ieee ieee NOUN bracis-33639 297 2 access access NOUN bracis-33639 297 3 11 11 NUM bracis-33639 297 4 , , PUNCT bracis-33639 297 5 60153–60170 60153–60170 NUM bracis-33639 297 6 ( ( PUNCT bracis-33639 297 7 2023 2023 NUM bracis-33639 297 8 ) ) PUNCT bracis-33639 297 9 article article NOUN bracis-33639 297 10     SPACE bracis-33639 297 11 google google PROPN bracis-33639 297 12 scholar scholar NOUN bracis-33639 297 13     SPACE bracis-33639 297 14 medsker medsker NOUN bracis-33639 297 15 , , PUNCT bracis-33639 297 16 l.r l.r PROPN bracis-33639 297 17 . . PROPN bracis-33639 297 18 , , PUNCT bracis-33639 297 19 jain jain PROPN bracis-33639 297 20 , , PUNCT bracis-33639 297 21 l. l. PROPN bracis-33639 297 22 , , PUNCT bracis-33639 297 23 et et PROPN bracis-33639 297 24 al al PROPN bracis-33639 297 25 . . PUNCT bracis-33639 297 26 : : PUNCT bracis-33639 298 1 recurrent recurrent ADJ bracis-33639 298 2 neural neural ADJ bracis-33639 298 3 networks network NOUN bracis-33639 298 4 . . PUNCT bracis-33639 299 1 des des PROPN bracis-33639 299 2 . . PROPN bracis-33639 299 3 appl appl PROPN bracis-33639 299 4 . . PUNCT bracis-33639 300 1 5(64–67 5(64–67 X bracis-33639 300 2 ) ) PUNCT bracis-33639 300 3 , , PUNCT bracis-33639 300 4 2 2 NUM bracis-33639 300 5 ( ( PUNCT bracis-33639 300 6 2001 2001 NUM bracis-33639 300 7 ) ) PUNCT bracis-33639 300 8 math math NOUN bracis-33639 300 9     SPACE bracis-33639 300 10 google google PROPN bracis-33639 300 11 scholar scholar NOUN bracis-33639 300 12     SPACE bracis-33639 300 13 oh oh INTJ bracis-33639 300 14 , , PUNCT bracis-33639 300 15 g. g. PROPN bracis-33639 300 16 , , PUNCT bracis-33639 300 17 song song NOUN bracis-33639 300 18 , , PUNCT bracis-33639 300 19 j. j. PROPN bracis-33639 300 20 , , PUNCT bracis-33639 300 21 park park PROPN bracis-33639 300 22 , , PUNCT bracis-33639 300 23 h. h. PROPN bracis-33639 300 24 , , PUNCT bracis-33639 300 25 na na PROPN bracis-33639 300 26 , , PUNCT bracis-33639 300 27 c. c. NOUN bracis-33639 300 28 : : PUNCT bracis-33639 300 29 evaluation evaluation NOUN bracis-33639 300 30 of of ADP bracis-33639 300 31 random random ADJ bracis-33639 300 32 forest forest NOUN bracis-33639 300 33 in in ADP bracis-33639 300 34 crime crime NOUN bracis-33639 300 35 prediction prediction NOUN bracis-33639 300 36 : : PUNCT bracis-33639 300 37 comparing compare VERB bracis-33639 300 38 three three NUM bracis-33639 300 39 - - PUNCT bracis-33639 300 40 layered layer VERB bracis-33639 300 41 random random ADJ bracis-33639 300 42 forest forest NOUN bracis-33639 300 43 and and CCONJ bracis-33639 300 44 logistic logistic ADJ bracis-33639 300 45 regression regression NOUN bracis-33639 300 46 . . PUNCT bracis-33639 301 1 deviant deviant ADJ bracis-33639 301 2 behav behav NOUN bracis-33639 301 3 . . PUNCT bracis-33639 302 1 43(9 43(9 NUM bracis-33639 302 2 ) ) PUNCT bracis-33639 302 3 , , PUNCT bracis-33639 302 4 1036–1049 1036–1049 NUM bracis-33639 302 5 ( ( PUNCT bracis-33639 302 6 2022 2022 NUM bracis-33639 302 7 ) ) PUNCT bracis-33639 302 8 article article NOUN bracis-33639 302 9     SPACE bracis-33639 302 10 math math NOUN bracis-33639 302 11     SPACE bracis-33639 302 12 google google PROPN bracis-33639 302 13 scholar scholar NOUN bracis-33639 302 14     SPACE bracis-33639 302 15 o’shea o’shea PROPN bracis-33639 302 16 , , PUNCT bracis-33639 302 17 k. k. PROPN bracis-33639 302 18 , , PUNCT bracis-33639 302 19 nash nash PROPN bracis-33639 302 20 , , PUNCT bracis-33639 302 21 r. r. PROPN bracis-33639 302 22 : : PUNCT bracis-33639 302 23 an an DET bracis-33639 302 24 introduction introduction NOUN bracis-33639 302 25 to to ADP bracis-33639 302 26 convolutional convolutional ADJ bracis-33639 302 27 neural neural ADJ bracis-33639 302 28 networks network NOUN bracis-33639 302 29 . . PUNCT bracis-33639 303 1 arxiv arxiv PROPN bracis-33639 303 2 preprint preprint VERB bracis-33639 303 3 arxiv:1511.08458 arxiv:1511.08458 NOUN bracis-33639 303 4 ( ( PUNCT bracis-33639 303 5 2015 2015 NUM bracis-33639 303 6 ) ) PUNCT bracis-33639 303 7 pareja pareja PROPN bracis-33639 303 8 , , PUNCT bracis-33639 303 9 a. a. PROPN bracis-33639 303 10 , , PUNCT bracis-33639 303 11 et et PROPN bracis-33639 303 12 al al PROPN bracis-33639 303 13 . . PUNCT bracis-33639 303 14 : : PUNCT bracis-33639 304 1 evolvegcn evolvegcn ADV bracis-33639 304 2 : : PUNCT bracis-33639 304 3 evolving evolve VERB bracis-33639 304 4 graph graph NOUN bracis-33639 304 5 convolutional convolutional ADJ bracis-33639 304 6 networks network NOUN bracis-33639 304 7 for for ADP bracis-33639 304 8 dynamic dynamic ADJ bracis-33639 304 9 graphs graph NOUN bracis-33639 304 10 . . PUNCT bracis-33639 305 1 arxiv arxiv PROPN bracis-33639 305 2 abs/1902.10191 abs/1902.10191 PROPN bracis-33639 305 3 ( ( PUNCT bracis-33639 305 4 2019 2019 NUM bracis-33639 305 5 ) ) PUNCT bracis-33639 305 6 . . PUNCT bracis-33639 306 1 https://api.semanticscholar.org/corpusid:67856459 https://api.semanticscholar.org/corpusid:67856459 PROPN bracis-33639 306 2 pedregosa pedregosa PROPN bracis-33639 306 3 , , PUNCT bracis-33639 306 4 f. f. PROPN bracis-33639 306 5 , , PUNCT bracis-33639 306 6 et et PROPN bracis-33639 306 7     SPACE bracis-33639 306 8 al al PROPN bracis-33639 306 9 . . PROPN bracis-33639 306 10 : : PUNCT bracis-33639 307 1 scikit scikit NOUN bracis-33639 307 2 - - PUNCT bracis-33639 307 3 learn learn VERB bracis-33639 307 4 : : PUNCT bracis-33639 307 5 machine machine NOUN bracis-33639 307 6 learning learning NOUN bracis-33639 307 7 in in ADP bracis-33639 307 8 python python PROPN bracis-33639 307 9 . . PUNCT bracis-33639 308 1 j. j. PROPN bracis-33639 308 2 mach mach PROPN bracis-33639 308 3 . . PUNCT bracis-33639 309 1 learn learn VERB bracis-33639 309 2 . . PUNCT bracis-33639 310 1 res re NOUN bracis-33639 310 2 . . PROPN bracis-33639 311 1 12 12 NUM bracis-33639 311 2 , , PUNCT bracis-33639 311 3 2825–2830 2825–2830 NUM bracis-33639 311 4 ( ( PUNCT bracis-33639 311 5 2011 2011 NUM bracis-33639 311 6 ) ) PUNCT bracis-33639 311 7 google google PROPN bracis-33639 311 8 scholar scholar NOUN bracis-33639 311 9     SPACE bracis-33639 311 10 geosampa geosampa NOUN bracis-33639 311 11 . . PUNCT bracis-33639 312 1 https://geosampa.prefeitura.sp.gov.br https://geosampa.prefeitura.sp.gov.br NOUN bracis-33639 312 2 . . PROPN bracis-33639 313 1 accessed access VERB bracis-33639 313 2 20 20 NUM bracis-33639 313 3 june june PROPN bracis-33639 313 4 2024 2024 NUM bracis-33639 313 5 salinas salina NOUN bracis-33639 313 6 , , PUNCT bracis-33639 313 7 k. k. PROPN bracis-33639 313 8 , , PUNCT bracis-33639 313 9 gonçalves gonçalves PROPN bracis-33639 313 10 , , PUNCT bracis-33639 313 11 t. t. PROPN bracis-33639 313 12 , , PUNCT bracis-33639 313 13 barella barella NOUN bracis-33639 313 14 , , PUNCT bracis-33639 313 15 v. v. ADV bracis-33639 313 16 , , PUNCT bracis-33639 313 17 vieira vieira NOUN bracis-33639 313 18 , , PUNCT bracis-33639 313 19 t. t. PROPN bracis-33639 313 20 , , PUNCT bracis-33639 313 21 nonato nonato PROPN bracis-33639 313 22 , , PUNCT bracis-33639 313 23 l.g l.g PROPN bracis-33639 313 24 . . PUNCT bracis-33639 313 25 : : PUNCT bracis-33639 314 1 cityhub cityhub PROPN bracis-33639 314 2 : : PUNCT bracis-33639 314 3 a a DET bracis-33639 314 4 library library NOUN bracis-33639 314 5 for for ADP bracis-33639 314 6 urban urban ADJ bracis-33639 314 7 data datum NOUN bracis-33639 314 8 integration integration NOUN bracis-33639 314 9 . . PUNCT bracis-33639 315 1 in in ADP bracis-33639 315 2 : : PUNCT bracis-33639 315 3 2022 2022 NUM bracis-33639 315 4 35th 35th ADJ bracis-33639 315 5 sibgrapi sibgrapi ADJ bracis-33639 315 6 conference conference NOUN bracis-33639 315 7 on on ADP bracis-33639 315 8 graphics graphic NOUN bracis-33639 315 9 , , PUNCT bracis-33639 315 10 patterns pattern NOUN bracis-33639 315 11 and and CCONJ bracis-33639 315 12 images image NOUN bracis-33639 315 13 ( ( PUNCT bracis-33639 315 14 sibgrapi sibgrapi ADJ bracis-33639 315 15 ) ) PUNCT bracis-33639 315 16 , , PUNCT bracis-33639 315 17 vol vol NOUN bracis-33639 315 18 . . PUNCT bracis-33639 315 19     SPACE bracis-33639 315 20 1 1 NUM bracis-33639 315 21 , , PUNCT bracis-33639 315 22 pp pp ADJ bracis-33639 315 23 . . PUNCT bracis-33639 315 24 43–48 43–48 NUM bracis-33639 315 25 . . PUNCT bracis-33639 315 26 ieee ieee NOUN bracis-33639 315 27 ( ( PUNCT bracis-33639 315 28 2022 2022 NUM bracis-33639 315 29 ) ) PUNCT bracis-33639 315 30 google google NOUN bracis-33639 315 31 scholar scholar NOUN bracis-33639 315 32     SPACE bracis-33639 315 33 sankar sankar NOUN bracis-33639 315 34 , , PUNCT bracis-33639 315 35 a. a. PROPN bracis-33639 315 36 , , PUNCT bracis-33639 315 37 wu wu PROPN bracis-33639 315 38 , , PUNCT bracis-33639 315 39 y. y. PROPN bracis-33639 315 40 , , PUNCT bracis-33639 315 41 gou gou PROPN bracis-33639 315 42 , , PUNCT bracis-33639 315 43 l. l. PROPN bracis-33639 315 44 , , PUNCT bracis-33639 315 45 zhang zhang PROPN bracis-33639 315 46 , , PUNCT bracis-33639 315 47 w. w. PROPN bracis-33639 315 48 , , PUNCT bracis-33639 315 49 yang yang PROPN bracis-33639 315 50 , , PUNCT bracis-33639 315 51 h. h. PROPN bracis-33639 315 52 : : PUNCT bracis-33639 315 53 dysat dysat VERB bracis-33639 315 54 : : PUNCT bracis-33639 315 55 deep deep ADJ bracis-33639 315 56 neural neural ADJ bracis-33639 315 57 representation representation NOUN bracis-33639 315 58 learning learn VERB bracis-33639 315 59 on on ADP bracis-33639 315 60 dynamic dynamic ADJ bracis-33639 315 61 graphs graph NOUN bracis-33639 315 62 via via ADP bracis-33639 315 63 self self NOUN bracis-33639 315 64 - - PUNCT bracis-33639 315 65 attention attention NOUN bracis-33639 315 66 networks network NOUN bracis-33639 315 67 . . PUNCT bracis-33639 316 1 in in ADP bracis-33639 316 2 : : PUNCT bracis-33639 316 3 proceedings proceeding NOUN bracis-33639 316 4 of of ADP bracis-33639 316 5 the the DET bracis-33639 316 6 13th 13th ADJ bracis-33639 316 7 international international ADJ bracis-33639 316 8 conference conference NOUN bracis-33639 316 9 on on ADP bracis-33639 316 10 web web NOUN bracis-33639 316 11 search search NOUN bracis-33639 316 12 and and CCONJ bracis-33639 316 13 data datum NOUN bracis-33639 316 14 mining mining NOUN bracis-33639 316 15 , , PUNCT bracis-33639 316 16 wsdm wsdm VERB bracis-33639 316 17 2020 2020 NUM bracis-33639 316 18 , , PUNCT bracis-33639 316 19 pp pp ADP bracis-33639 316 20 519–527 519–527 NUM bracis-33639 316 21 . . PUNCT bracis-33639 317 1 association association NOUN bracis-33639 317 2 for for ADP bracis-33639 317 3 computing compute VERB bracis-33639 317 4 machinery machinery NOUN bracis-33639 317 5 , , PUNCT bracis-33639 317 6 new new PROPN bracis-33639 317 7 york york PROPN bracis-33639 317 8 ( ( PUNCT bracis-33639 317 9 2020 2020 NUM bracis-33639 317 10 ) ) PUNCT bracis-33639 317 11 . . PUNCT bracis-33639 318 1 https://doi.org/10.1145/3336191.3371845 https://doi.org/10.1145/3336191.3371845 PROPN bracis-33639 319 1 secretaria secretaria PROPN bracis-33639 319 2 de de PROPN bracis-33639 319 3 segurança segurança PROPN bracis-33639 319 4 pública pública PROPN bracis-33639 319 5 . . PUNCT bracis-33639 320 1 https://www.ssp.sp.gov.br/estatistica/consultas https://www.ssp.sp.gov.br/estatistica/consulta NOUN bracis-33639 320 2 . . PUNCT bracis-33639 320 3 accessed access VERB bracis-33639 320 4 20 20 NUM bracis-33639 320 5 june june PROPN bracis-33639 320 6 2024 2024 NUM bracis-33639 320 7 shi shi PROPN bracis-33639 320 8 , , PUNCT bracis-33639 320 9 x. x. PROPN bracis-33639 320 10 , , PUNCT bracis-33639 320 11 chen chen PROPN bracis-33639 320 12 , , PUNCT bracis-33639 320 13 z. z. PROPN bracis-33639 320 14 , , PUNCT bracis-33639 320 15 wang wang PROPN bracis-33639 320 16 , , PUNCT bracis-33639 320 17 h. h. PROPN bracis-33639 320 18 , , PUNCT bracis-33639 320 19 yeung yeung PROPN bracis-33639 320 20 , , PUNCT bracis-33639 320 21 d.y d.y PROPN bracis-33639 320 22 . . PROPN bracis-33639 320 23 , , PUNCT bracis-33639 320 24 wong wong PROPN bracis-33639 320 25 , , PUNCT bracis-33639 320 26 w.k w.k PROPN bracis-33639 320 27 . . PROPN bracis-33639 320 28 , , PUNCT bracis-33639 320 29 woo woo PROPN bracis-33639 320 30 , , PUNCT bracis-33639 320 31 w.c w.c PROPN bracis-33639 320 32 . . PROPN bracis-33639 320 33 : : PUNCT bracis-33639 320 34 convolutional convolutional ADJ bracis-33639 320 35 lstm lstm ADJ bracis-33639 320 36 network network NOUN bracis-33639 320 37 : : PUNCT bracis-33639 320 38 a a DET bracis-33639 320 39 machine machine NOUN bracis-33639 320 40 learning learn VERB bracis-33639 320 41 approach approach NOUN bracis-33639 320 42 for for ADP bracis-33639 320 43 precipitation precipitation NOUN bracis-33639 320 44 nowcasting nowcasting NOUN bracis-33639 320 45 . . PUNCT bracis-33639 321 1 in in ADP bracis-33639 321 2 : : PUNCT bracis-33639 321 3 cortes corte NOUN bracis-33639 321 4 , , PUNCT bracis-33639 321 5 c. c. PROPN bracis-33639 321 6 , , PUNCT bracis-33639 321 7 lawrence lawrence PROPN bracis-33639 321 8 , , PUNCT bracis-33639 321 9 n. n. PROPN bracis-33639 321 10 , , PUNCT bracis-33639 321 11 lee lee PROPN bracis-33639 321 12 , , PUNCT bracis-33639 321 13 d. d. PROPN bracis-33639 321 14 , , PUNCT bracis-33639 321 15 sugiyama sugiyama NOUN bracis-33639 321 16 , , PUNCT bracis-33639 321 17 m. m. NOUN bracis-33639 321 18 , , PUNCT bracis-33639 321 19 garnett garnett PROPN bracis-33639 321 20 , , PUNCT bracis-33639 321 21 r. r. PROPN bracis-33639 321 22 ( ( PUNCT bracis-33639 321 23 eds eds PROPN bracis-33639 321 24 . . PUNCT bracis-33639 321 25 ) ) PUNCT bracis-33639 321 26 advances advance NOUN bracis-33639 321 27 in in ADP bracis-33639 321 28 neural neural ADJ bracis-33639 321 29 information information NOUN bracis-33639 321 30 processing processing NOUN bracis-33639 321 31 systems system NOUN bracis-33639 321 32 , , PUNCT bracis-33639 321 33 vol vol NOUN bracis-33639 321 34 . . PUNCT bracis-33639 321 35     SPACE bracis-33639 321 36 28 28 NUM bracis-33639 321 37 . . PUNCT bracis-33639 322 1 curran curran PROPN bracis-33639 322 2 associates associates PROPN bracis-33639 322 3 , , PUNCT bracis-33639 322 4 inc inc PROPN bracis-33639 322 5 . . PROPN bracis-33639 322 6 ( ( PUNCT bracis-33639 322 7 2015 2015 NUM bracis-33639 322 8 ) ) PUNCT bracis-33639 322 9 . . PUNCT bracis-33639 323 1 https://proceedings.neurips.cc/paper_files/paper/2015/file/07563a3fe3bbe7e3ba84431ad9d055af-paper.pdf https://proceedings.neurips.cc/paper_files/paper/2015/file/07563a3fe3bbe7e3ba84431ad9d055af-paper.pdf PROPN bracis-33639 323 2 thomas thomas PROPN bracis-33639 323 3 , , PUNCT bracis-33639 323 4 a. a. PROPN bracis-33639 323 5 , , PUNCT bracis-33639 323 6 sobhana sobhana PROPN bracis-33639 323 7 , , PUNCT bracis-33639 323 8 n. n. NOUN bracis-33639 323 9 : : PUNCT bracis-33639 323 10 a a DET bracis-33639 323 11 survey survey NOUN bracis-33639 323 12 on on ADP bracis-33639 323 13 crime crime NOUN bracis-33639 323 14 analysis analysis NOUN bracis-33639 323 15 and and CCONJ bracis-33639 323 16 prediction prediction NOUN bracis-33639 323 17 . . PUNCT bracis-33639 324 1 mater mater NOUN bracis-33639 324 2 . . PUNCT bracis-33639 325 1 today today NOUN bracis-33639 325 2 proc proc NOUN bracis-33639 325 3 . . PUNCT bracis-33639 326 1 58 58 NUM bracis-33639 326 2 , , PUNCT bracis-33639 326 3 310–315 310–315 NUM bracis-33639 326 4 ( ( PUNCT bracis-33639 326 5 2022 2022 NUM bracis-33639 326 6 ) ) PUNCT bracis-33639 326 7 article article NOUN bracis-33639 326 8     SPACE bracis-33639 326 9 math math NOUN bracis-33639 326 10     SPACE bracis-33639 326 11 google google PROPN bracis-33639 326 12 scholar scholar NOUN bracis-33639 326 13     SPACE bracis-33639 326 14 wu wu PROPN bracis-33639 326 15 , , PUNCT bracis-33639 326 16 s. s. PROPN bracis-33639 326 17 , , PUNCT bracis-33639 326 18 sun sun PROPN bracis-33639 326 19 , , PUNCT bracis-33639 326 20 f. f. PROPN bracis-33639 326 21 , , PUNCT bracis-33639 326 22 zhang zhang PROPN bracis-33639 326 23 , , PUNCT bracis-33639 326 24 w. w. PROPN bracis-33639 326 25 , , PUNCT bracis-33639 326 26 xie xie PROPN bracis-33639 326 27 , , PUNCT bracis-33639 326 28 x. x. PROPN bracis-33639 326 29 , , PUNCT bracis-33639 326 30 cui cui PROPN bracis-33639 326 31 , , PUNCT bracis-33639 326 32 b. b. PROPN bracis-33639 326 33 : : PUNCT bracis-33639 326 34 graph graph VERB bracis-33639 326 35 neural neural ADJ bracis-33639 326 36 networks network NOUN bracis-33639 326 37 in in ADP bracis-33639 326 38 recommender recommender NOUN bracis-33639 326 39 systems system NOUN bracis-33639 326 40 : : PUNCT bracis-33639 326 41 a a DET bracis-33639 326 42 survey survey NOUN bracis-33639 326 43 . . PUNCT bracis-33639 327 1 acm acm PROPN bracis-33639 327 2 comput comput NOUN bracis-33639 327 3 . . PUNCT bracis-33639 328 1 surv surv NOUN bracis-33639 328 2 . . PUNCT bracis-33639 329 1 55(5 55(5 NUM bracis-33639 329 2 ) ) PUNCT bracis-33639 329 3 ( ( PUNCT bracis-33639 329 4 2022 2022 NUM bracis-33639 329 5 ) ) PUNCT bracis-33639 329 6 . . PUNCT bracis-33639 330 1 https://doi.org/10.1145/3535101 https://doi.org/10.1145/3535101 PROPN bracis-33639 330 2 zhou zhou PROPN bracis-33639 330 3 , , PUNCT bracis-33639 330 4 j. j. PROPN bracis-33639 330 5 , , PUNCT bracis-33639 330 6 et et PROPN bracis-33639 330 7 al al PROPN bracis-33639 330 8 . . PUNCT bracis-33639 330 9 : : PUNCT bracis-33639 331 1 graph graph VERB bracis-33639 331 2 neural neural ADJ bracis-33639 331 3 networks network NOUN bracis-33639 331 4 : : PUNCT bracis-33639 331 5 a a DET bracis-33639 331 6 review review NOUN bracis-33639 331 7 of of ADP bracis-33639 331 8 methods method NOUN bracis-33639 331 9 and and CCONJ bracis-33639 331 10 applications application NOUN bracis-33639 331 11 . . PUNCT bracis-33639 332 1 ai ai VERB bracis-33639 332 2 open open ADJ bracis-33639 332 3 1 1 NUM bracis-33639 332 4 , , PUNCT bracis-33639 332 5 57–81 57–81 NUM bracis-33639 332 6 ( ( PUNCT bracis-33639 332 7 2020 2020 NUM bracis-33639 332 8 ) ) PUNCT bracis-33639 332 9 . . PUNCT bracis-33639 333 1 https://doi.org/10.1016/j.aiopen.2021.01.001 https://doi.org/10.1016/j.aiopen.2021.01.001 PROPN bracis-33639 333 2 article article NOUN bracis-33639 333 3     SPACE bracis-33639 333 4 math math NOUN bracis-33639 333 5     SPACE bracis-33639 333 6 google google PROPN bracis-33639 333 7 scholar scholar NOUN bracis-33639 333 8     SPACE bracis-33639 333 9 download download NOUN bracis-33639 333 10 references reference NOUN bracis-33639 333 11 acknowledgments acknowledgment NOUN bracis-33639 333 12 this this DET bracis-33639 333 13 work work NOUN bracis-33639 333 14 was be AUX bracis-33639 333 15 supported support VERB bracis-33639 333 16 by by ADP bracis-33639 333 17 cnpq cnpq NOUN bracis-33639 333 18 ( ( PUNCT bracis-33639 333 19 # # SYM bracis-33639 333 20 307184/2021 307184/2021 NUM bracis-33639 333 21 - - PUNCT bracis-33639 333 22 8) 8) NUM bracis-33639 333 23 and and CCONJ bracis-33639 333 24 fapesp fapesp NOUN bracis-33639 333 25 ( ( PUNCT bracis-33639 333 26 # # SYM bracis-33639 333 27 2022/09091 2022/09091 NUM bracis-33639 333 28 - - SYM bracis-33639 333 29 8 8 NUM bracis-33639 333 30 , , PUNCT bracis-33639 333 31 # # SYM bracis-33639 333 32 2023/16334 2023/16334 NUM bracis-33639 333 33 - - SYM bracis-33639 333 34 7 7 NUM bracis-33639 333 35 , , PUNCT bracis-33639 333 36 # # SYM bracis-33639 333 37 2013/07375 2013/07375 NUM bracis-33639 333 38 - - SYM bracis-33639 333 39 0 0 NUM bracis-33639 333 40 , , PUNCT bracis-33639 333 41 # # SYM bracis-33639 333 42 2023/15618 2023/15618 NUM bracis-33639 333 43 - - SYM bracis-33639 333 44 1 1 NUM bracis-33639 333 45 ) ) PUNCT bracis-33639 333 46 . . PUNCT bracis-33639 334 1 the the DET bracis-33639 334 2 opinions opinion NOUN bracis-33639 334 3 , , PUNCT bracis-33639 334 4 hypotheses hypothesis NOUN bracis-33639 334 5 , , PUNCT bracis-33639 334 6 conclusions conclusion NOUN bracis-33639 334 7 and and CCONJ bracis-33639 334 8 recommendations recommendation NOUN bracis-33639 334 9 expressed express VERB bracis-33639 334 10 in in ADP bracis-33639 334 11 this this DET bracis-33639 334 12 material material NOUN bracis-33639 334 13 are be AUX bracis-33639 334 14 responsibility responsibility NOUN bracis-33639 334 15 of of ADP bracis-33639 334 16 the the DET bracis-33639 334 17 author(s author(s NOUN bracis-33639 334 18 ) ) PUNCT bracis-33639 334 19 and and CCONJ bracis-33639 334 20 do do AUX bracis-33639 334 21 not not PART bracis-33639 334 22 necessarily necessarily ADV bracis-33639 334 23 reflect reflect VERB bracis-33639 334 24 cnpq cnpq NOUN bracis-33639 334 25 and and CCONJ bracis-33639 334 26 fapesp fapesp ADJ bracis-33639 334 27 views view NOUN bracis-33639 334 28 . . PUNCT bracis-33639 335 1 author author NOUN bracis-33639 335 2 information information NOUN bracis-33639 335 3 authors author NOUN bracis-33639 335 4 and and CCONJ bracis-33639 335 5 affiliations affiliation NOUN bracis-33639 335 6 university university PROPN bracis-33639 335 7 of of ADP bracis-33639 335 8 são são PROPN bracis-33639 335 9 paulo paulo PROPN bracis-33639 335 10 , , PUNCT bracis-33639 335 11 são são PROPN bracis-33639 335 12 carlos carlos PROPN bracis-33639 335 13 , , PUNCT bracis-33639 335 14 brazil brazil PROPN bracis-33639 335 15 waqar waqar PROPN bracis-33639 335 16 hassan hassan PROPN bracis-33639 335 17 , , PUNCT bracis-33639 335 18     SPACE bracis-33639 335 19 marvin marvin PROPN bracis-33639 335 20 mendes mendes PROPN bracis-33639 335 21 cabral cabral PROPN bracis-33639 335 22 , , PUNCT bracis-33639 335 23     SPACE bracis-33639 335 24 antonio antonio PROPN bracis-33639 335 25 castelo castelo PROPN bracis-33639 335 26 filho filho PROPN bracis-33639 335 27     SPACE bracis-33639 335 28 & & CCONJ bracis-33639 335 29     SPACE bracis-33639 335 30 luis luis PROPN bracis-33639 335 31 gustavo gustavo PROPN bracis-33639 335 32 nonato nonato PROPN bracis-33639 335 33 federal federal PROPN bracis-33639 335 34 university university PROPN bracis-33639 335 35 of of ADP bracis-33639 335 36 são são PROPN bracis-33639 335 37 carlos carlos PROPN bracis-33639 335 38 , , PUNCT bracis-33639 335 39 são são PROPN bracis-33639 335 40 carlos carlos PROPN bracis-33639 335 41 , , PUNCT bracis-33639 335 42 brazil brazil PROPN bracis-33639 335 43 thiago thiago PROPN bracis-33639 335 44 rodrigo rodrigo PROPN bracis-33639 335 45 ramos ramos PROPN bracis-33639 335 46 authors author NOUN bracis-33639 335 47 waqar waqar VERB bracis-33639 335 48 hassanview hassanview PROPN bracis-33639 335 49 author author NOUN bracis-33639 335 50 publications publication NOUN bracis-33639 335 51 search search NOUN bracis-33639 335 52 author author NOUN bracis-33639 335 53 on on ADP bracis-33639 335 54 : : PUNCT bracis-33639 335 55 pubmed pubmed PROPN bracis-33639 335 56     SPACE bracis-33639 335 57 google google PROPN bracis-33639 335 58 scholar scholar PROPN bracis-33639 335 59 marvin marvin PROPN bracis-33639 335 60 mendes mendes PROPN bracis-33639 335 61 cabralview cabralview VERB bracis-33639 335 62 author author NOUN bracis-33639 335 63 publications publication NOUN bracis-33639 335 64 search search NOUN bracis-33639 335 65 author author NOUN bracis-33639 335 66 on on ADP bracis-33639 335 67 : : PUNCT bracis-33639 335 68 pubmed pubmed PROPN bracis-33639 335 69     SPACE bracis-33639 335 70 google google PROPN bracis-33639 335 71 scholar scholar PROPN bracis-33639 335 72 thiago thiago PROPN bracis-33639 335 73 rodrigo rodrigo PROPN bracis-33639 335 74 ramosview ramosview PROPN bracis-33639 335 75 author author NOUN bracis-33639 335 76 publications publication VERB bracis-33639 335 77 search search NOUN bracis-33639 335 78 author author NOUN bracis-33639 335 79 on on ADP bracis-33639 335 80 : : PUNCT bracis-33639 335 81 pubmed pubmed PROPN bracis-33639 335 82     SPACE bracis-33639 335 83 google google PROPN bracis-33639 335 84 scholar scholar PROPN bracis-33639 335 85 antonio antonio PROPN bracis-33639 335 86 castelo castelo PROPN bracis-33639 335 87 filhoview filhoview ADJ bracis-33639 335 88 author author NOUN bracis-33639 335 89 publications publication NOUN bracis-33639 335 90 search search NOUN bracis-33639 335 91 author author NOUN bracis-33639 335 92 on on ADP bracis-33639 335 93 : : PUNCT bracis-33639 335 94 pubmed pubmed PROPN bracis-33639 335 95     SPACE bracis-33639 335 96 google google PROPN bracis-33639 335 97 scholar scholar PROPN bracis-33639 335 98 luis luis PROPN bracis-33639 335 99 gustavo gustavo PROPN bracis-33639 335 100 nonatoview nonatoview NOUN bracis-33639 335 101 author author NOUN bracis-33639 335 102 publications publication NOUN bracis-33639 335 103 search search NOUN bracis-33639 335 104 author author NOUN bracis-33639 335 105 on on ADP bracis-33639 335 106 : : PUNCT bracis-33639 335 107 pubmed pubmed PROPN bracis-33639 335 108     SPACE bracis-33639 335 109 google google PROPN bracis-33639 335 110 scholar scholar NOUN bracis-33639 335 111 corresponding correspond VERB bracis-33639 335 112 author author NOUN bracis-33639 335 113 correspondence correspondence NOUN bracis-33639 335 114 to to ADP bracis-33639 335 115 waqar waqar PROPN bracis-33639 335 116 hassan hassan PROPN bracis-33639 335 117 . . PUNCT bracis-33639 336 1 editor editor NOUN bracis-33639 336 2 information information NOUN bracis-33639 336 3 editors editor NOUN bracis-33639 336 4 and and CCONJ bracis-33639 336 5 affiliations affiliation NOUN bracis-33639 336 6 universidade universidade VERB bracis-33639 336 7 federal federal PROPN bracis-33639 336 8 fluminense fluminense PROPN bracis-33639 336 9 , , PUNCT bracis-33639 336 10 niterói niterói PROPN bracis-33639 336 11 , , PUNCT bracis-33639 336 12 brazil brazil PROPN bracis-33639 336 13 aline aline PROPN bracis-33639 336 14 paes paes PROPN bracis-33639 336 15 instituto instituto PROPN bracis-33639 336 16 tecnológico tecnológico PROPN bracis-33639 336 17 de de PROPN bracis-33639 336 18 aeronáutica aeronáutica PROPN bracis-33639 336 19 , , PUNCT bracis-33639 336 20 são são PROPN bracis-33639 336 21 josé josé PROPN bracis-33639 336 22 dos dos PROPN bracis-33639 336 23 campos campos PROPN bracis-33639 336 24 , , PUNCT bracis-33639 336 25 brazil brazil PROPN bracis-33639 336 26 filipe filipe PROPN bracis-33639 336 27 a. a. PROPN bracis-33639 336 28 n. n. PROPN bracis-33639 336 29 verri verri PROPN bracis-33639 336 30 ethics ethic NOUN bracis-33639 336 31 declarations declaration VERB bracis-33639 336 32 disclosure disclosure NOUN bracis-33639 336 33 of of ADP bracis-33639 336 34 interests interest NOUN bracis-33639 336 35 the the DET bracis-33639 336 36 authors author NOUN bracis-33639 336 37 have have VERB bracis-33639 336 38 no no DET bracis-33639 336 39 competing compete VERB bracis-33639 336 40 interests interest NOUN bracis-33639 336 41 to to PART bracis-33639 336 42 declare declare VERB bracis-33639 336 43 that that PRON bracis-33639 336 44 are be AUX bracis-33639 336 45 relevant relevant ADJ bracis-33639 336 46 to to ADP bracis-33639 336 47 the the DET bracis-33639 336 48 content content NOUN bracis-33639 336 49 of of ADP bracis-33639 336 50 this this DET bracis-33639 336 51 article article NOUN bracis-33639 336 52 . . PUNCT bracis-33639 337 1 rights right NOUN bracis-33639 337 2 and and CCONJ bracis-33639 337 3 permissions permission NOUN bracis-33639 337 4 reprints reprint NOUN bracis-33639 337 5 and and CCONJ bracis-33639 337 6 permissions permission VERB bracis-33639 337 7 copyright copyright NOUN bracis-33639 337 8 information information NOUN bracis-33639 337 9 © © ADP bracis-33639 337 10 2025 2025 NUM bracis-33639 337 11 the the DET bracis-33639 337 12 author(s author(s NOUN bracis-33639 337 13 ) ) PUNCT bracis-33639 337 14 , , PUNCT bracis-33639 337 15 under under ADP bracis-33639 337 16 exclusive exclusive ADJ bracis-33639 337 17 license license NOUN bracis-33639 337 18 to to ADP bracis-33639 337 19 springer springer NOUN bracis-33639 337 20 nature nature PROPN bracis-33639 337 21 switzerland switzerland PROPN bracis-33639 337 22 ag ag PROPN bracis-33639 337 23 about about ADP bracis-33639 337 24 this this DET bracis-33639 337 25 paper paper NOUN bracis-33639 337 26 cite cite VERB bracis-33639 337 27 this this DET bracis-33639 337 28 paper paper NOUN bracis-33639 337 29 hassan hassan PROPN bracis-33639 337 30 , , PUNCT bracis-33639 337 31 w. w. PROPN bracis-33639 337 32 , , PUNCT bracis-33639 337 33 cabral cabral PROPN bracis-33639 337 34 , , PUNCT bracis-33639 337 35 m.m m.m PROPN bracis-33639 337 36 . . PROPN bracis-33639 337 37 , , PUNCT bracis-33639 337 38 ramos ramos PROPN bracis-33639 337 39 , , PUNCT bracis-33639 337 40 t.r t.r PROPN bracis-33639 337 41 . . PROPN bracis-33639 337 42 , , PUNCT bracis-33639 337 43 filho filho PROPN bracis-33639 337 44 , , PUNCT bracis-33639 337 45 a.c a.c PROPN bracis-33639 337 46 . . PROPN bracis-33639 337 47 , , PUNCT bracis-33639 337 48 nonato nonato PROPN bracis-33639 337 49 , , PUNCT bracis-33639 337 50 l.g l.g PROPN bracis-33639 337 51 . . PROPN bracis-33639 337 52 ( ( PUNCT bracis-33639 337 53 2025 2025 NUM bracis-33639 337 54 ) ) PUNCT bracis-33639 337 55 . . PUNCT bracis-33639 338 1 modeling modeling NOUN bracis-33639 338 2 and and CCONJ bracis-33639 338 3     SPACE bracis-33639 338 4 predicting predict VERB bracis-33639 338 5 crimes crime NOUN bracis-33639 338 6 in in ADP bracis-33639 338 7     SPACE bracis-33639 338 8 the the DET bracis-33639 338 9     SPACE bracis-33639 338 10 city city NOUN bracis-33639 338 11 of of ADP bracis-33639 338 12     SPACE bracis-33639 338 13 são são PROPN bracis-33639 338 14 paulo paulo NOUN bracis-33639 338 15 using use VERB bracis-33639 338 16 graph graph NOUN bracis-33639 338 17 neural neural ADJ bracis-33639 338 18 networks network NOUN bracis-33639 338 19 . . PUNCT bracis-33639 339 1 in in ADP bracis-33639 339 2 : : PUNCT bracis-33639 339 3 paes paes PROPN bracis-33639 339 4 , , PUNCT bracis-33639 339 5 a. a. PROPN bracis-33639 339 6 , , PUNCT bracis-33639 339 7 verri verri NOUN bracis-33639 339 8 , , PUNCT bracis-33639 339 9 f.a.n f.a.n ADJ bracis-33639 339 10 . . PUNCT bracis-33639 339 11 ( ( PUNCT bracis-33639 339 12 eds ed NOUN bracis-33639 339 13 ) ) PUNCT bracis-33639 339 14 intelligent intelligent ADJ bracis-33639 339 15 systems system NOUN bracis-33639 339 16 . . PUNCT bracis-33639 339 17 bracis bracis PROPN bracis-33639 339 18 2024 2024 NUM bracis-33639 339 19 . . PUNCT bracis-33639 340 1 lecture lecture NOUN bracis-33639 340 2 notes note NOUN bracis-33639 340 3 in in ADP bracis-33639 340 4 computer computer NOUN bracis-33639 340 5 science science NOUN bracis-33639 340 6 ( ( PUNCT bracis-33639 340 7 ) ) PUNCT bracis-33639 340 8 , , PUNCT bracis-33639 340 9 vol vol NOUN bracis-33639 340 10 15414 15414 NUM bracis-33639 340 11 . . PUNCT bracis-33639 341 1 springer springer NOUN bracis-33639 341 2 , , PUNCT bracis-33639 341 3 cham cham PROPN bracis-33639 341 4 . . PUNCT bracis-33639 342 1 https://doi.org/10.1007/978-3-031-79035-5_26 https://doi.org/10.1007/978-3-031-79035-5_26 PART bracis-33639 342 2 download download NOUN bracis-33639 342 3 citation citation NOUN bracis-33639 342 4 .ris .ris PUNCT bracis-33639 343 1 .enw .enw PROPN bracis-33639 343 2 .bib .bib PUNCT bracis-33639 344 1 doi doi PROPN bracis-33639 344 2 : : PUNCT bracis-33639 344 3 https://doi.org/10.1007/978-3-031-79035-5_26 https://doi.org/10.1007/978-3-031-79035-5_26 VERB bracis-33639 344 4 published publish VERB bracis-33639 344 5 : : PUNCT bracis-33639 344 6 30 30 NUM bracis-33639 344 7 january january PROPN bracis-33639 344 8 2025 2025 NUM bracis-33639 344 9 publisher publisher NOUN bracis-33639 344 10 name name NOUN bracis-33639 344 11 : : PUNCT bracis-33639 344 12 springer springer NOUN bracis-33639 344 13 , , PUNCT bracis-33639 344 14 cham cham PROPN bracis-33639 344 15 print print PROPN bracis-33639 344 16 isbn isbn PROPN bracis-33639 344 17 : : PUNCT bracis-33639 344 18 978 978 NUM bracis-33639 344 19 - - SYM bracis-33639 344 20 3 3 NUM bracis-33639 344 21 - - PUNCT bracis-33639 344 22 031 031 NUM bracis-33639 344 23 - - PUNCT bracis-33639 344 24 79034 79034 NUM bracis-33639 344 25 - - SYM bracis-33639 344 26 8 8 NUM bracis-33639 344 27 online online ADJ bracis-33639 344 28 isbn isbn NOUN bracis-33639 344 29 : : PUNCT bracis-33639 344 30 978 978 NUM bracis-33639 344 31 - - SYM bracis-33639 344 32 3 3 NUM bracis-33639 344 33 - - PUNCT bracis-33639 344 34 031 031 NUM bracis-33639 344 35 - - PUNCT bracis-33639 344 36 79035 79035 NUM bracis-33639 344 37 - - PUNCT bracis-33639 344 38 5 5 NUM bracis-33639 344 39 ebook ebook NOUN bracis-33639 344 40 packages package NOUN bracis-33639 344 41 : : PUNCT bracis-33639 344 42 computer computer NOUN bracis-33639 344 43 sciencecomputer sciencecomputer NOUN bracis-33639 344 44 science science NOUN bracis-33639 344 45 ( ( PUNCT bracis-33639 344 46 r0 r0 NOUN bracis-33639 344 47 ) ) PUNCT bracis-33639 344 48 share share VERB bracis-33639 344 49 this this DET bracis-33639 344 50 paper paper NOUN bracis-33639 344 51 anyone anyone PRON bracis-33639 344 52 you you PRON bracis-33639 344 53 share share VERB bracis-33639 344 54 the the DET bracis-33639 344 55 following follow VERB bracis-33639 344 56 link link NOUN bracis-33639 344 57 with with ADP bracis-33639 344 58 will will AUX bracis-33639 344 59 be be AUX bracis-33639 344 60 able able ADJ bracis-33639 344 61 to to PART bracis-33639 344 62 read read VERB bracis-33639 344 63 this this DET bracis-33639 344 64 content content NOUN bracis-33639 344 65 : : PUNCT bracis-33639 344 66 get get VERB bracis-33639 344 67 shareable shareable ADJ bracis-33639 344 68 linksorry linksorry NOUN bracis-33639 344 69 , , PUNCT bracis-33639 344 70 a a DET bracis-33639 344 71 shareable shareable ADJ bracis-33639 344 72 link link NOUN bracis-33639 344 73 is be AUX bracis-33639 344 74 not not PART bracis-33639 344 75 currently currently ADV bracis-33639 344 76 available available ADJ bracis-33639 344 77 for for ADP bracis-33639 344 78 this this DET bracis-33639 344 79 article article NOUN bracis-33639 344 80 . . PUNCT bracis-33639 345 1 copy copy VERB bracis-33639 345 2 shareable shareable ADJ bracis-33639 345 3 link link NOUN bracis-33639 345 4 to to PART bracis-33639 345 5 clipboard clipboard NOUN bracis-33639 345 6 provided provide VERB bracis-33639 345 7 by by ADP bracis-33639 345 8 the the DET bracis-33639 345 9 springer springer NOUN bracis-33639 345 10 nature nature PROPN bracis-33639 345 11 sharedit sharedit PROPN bracis-33639 345 12 content content NOUN bracis-33639 345 13 - - PUNCT bracis-33639 345 14 sharing share VERB bracis-33639 345 15 initiative initiative NOUN bracis-33639 345 16 keywords keyword NOUN bracis-33639 345 17 data datum NOUN bracis-33639 345 18 modeling model VERB bracis-33639 345 19 graph graph NOUN bracis-33639 345 20 neural neural ADJ bracis-33639 345 21 networks network NOUN bracis-33639 345 22 crime crime NOUN bracis-33639 345 23 prediction prediction NOUN bracis-33639 345 24 publish publish VERB bracis-33639 345 25 with with ADP bracis-33639 345 26 us us PROPN bracis-33639 345 27 policies policy NOUN bracis-33639 345 28 and and CCONJ bracis-33639 345 29 ethics ethic NOUN bracis-33639 345 30 search search NOUN bracis-33639 345 31 search search NOUN bracis-33639 345 32 by by ADP bracis-33639 345 33 keyword keyword NOUN bracis-33639 345 34 or or CCONJ bracis-33639 345 35 author author NOUN bracis-33639 345 36 search search NOUN bracis-33639 345 37 navigation navigation NOUN bracis-33639 345 38 find find VERB bracis-33639 345 39 a a DET bracis-33639 345 40 journal journal NOUN bracis-33639 345 41 publish publish VERB bracis-33639 345 42 with with ADP bracis-33639 345 43 us we PRON bracis-33639 345 44 track track VERB bracis-33639 345 45 your your PRON bracis-33639 345 46 research research NOUN bracis-33639 345 47 discover discover VERB bracis-33639 345 48 content content NOUN bracis-33639 345 49 journals journal NOUN bracis-33639 345 50 a a DET bracis-33639 345 51 - - PUNCT bracis-33639 345 52 z z NOUN bracis-33639 345 53 books book NOUN bracis-33639 345 54 a a DET bracis-33639 345 55 - - PUNCT bracis-33639 345 56 z z NOUN bracis-33639 345 57 publish publish NOUN bracis-33639 345 58 with with ADP bracis-33639 345 59 us us PROPN bracis-33639 345 60 journal journal PROPN bracis-33639 345 61 finder finder PROPN bracis-33639 345 62 publish publish VERB bracis-33639 345 63 your your PRON bracis-33639 345 64 research research NOUN bracis-33639 345 65 language language NOUN bracis-33639 345 66 editing edit VERB bracis-33639 345 67 open open ADJ bracis-33639 345 68 access access NOUN bracis-33639 345 69 publishing publishing NOUN bracis-33639 345 70 products product NOUN bracis-33639 345 71 and and CCONJ bracis-33639 345 72 services service NOUN bracis-33639 345 73 our our PRON bracis-33639 345 74 products product NOUN bracis-33639 345 75 librarians librarian VERB bracis-33639 345 76 societies society NOUN bracis-33639 345 77 partners partner NOUN bracis-33639 345 78 and and CCONJ bracis-33639 345 79 advertisers advertiser NOUN bracis-33639 345 80 our our PRON bracis-33639 345 81 brands brand NOUN bracis-33639 345 82 springer springer NOUN bracis-33639 345 83 nature nature PROPN bracis-33639 345 84 portfolio portfolio PROPN bracis-33639 345 85 bmc bmc PROPN bracis-33639 345 86 palgrave palgrave PROPN bracis-33639 345 87 macmillan macmillan PROPN bracis-33639 345 88 apress apress PROPN bracis-33639 345 89 discover discover VERB bracis-33639 345 90 your your PRON bracis-33639 345 91 privacy privacy NOUN bracis-33639 345 92 choices choice NOUN bracis-33639 345 93 / / SYM bracis-33639 345 94 manage manage NOUN bracis-33639 345 95 cookies cookie NOUN bracis-33639 345 96 your your PRON bracis-33639 345 97 us us PROPN bracis-33639 346 1 state state NOUN bracis-33639 346 2 privacy privacy NOUN bracis-33639 346 3 rights right NOUN bracis-33639 346 4 accessibility accessibility NOUN bracis-33639 346 5 statement statement NOUN bracis-33639 346 6 terms term NOUN bracis-33639 346 7 and and CCONJ bracis-33639 346 8 conditions condition NOUN bracis-33639 346 9 privacy privacy NOUN bracis-33639 346 10 policy policy NOUN bracis-33639 346 11 help help NOUN bracis-33639 346 12 and and CCONJ bracis-33639 346 13 support support VERB bracis-33639 346 14 legal legal ADJ bracis-33639 346 15 notice notice NOUN bracis-33639 346 16 cancel cancel VERB bracis-33639 346 17 contracts contract NOUN bracis-33639 346 18 here here ADV bracis-33639 346 19 129.74.145.123 129.74.145.123 NUM bracis-33639 346 20 hesburgh hesburgh PROPN bracis-33639 346 21 library library PROPN bracis-33639 346 22 er er INTJ bracis-33639 346 23 unit unit NOUN bracis-33639 346 24 ( ( PUNCT bracis-33639 346 25 3005732405 3005732405 NUM bracis-33639 346 26 ) ) PUNCT bracis-33639 346 27 northeast northeast ADJ bracis-33639 346 28 research research NOUN bracis-33639 346 29 libraries library NOUN bracis-33639 346 30 ( ( PUNCT bracis-33639 346 31 nerl nerl PROPN bracis-33639 346 32 ) ) PUNCT bracis-33639 346 33 ( ( PUNCT bracis-33639 346 34 8200828607 8200828607 NUM bracis-33639 346 35 ) ) PUNCT bracis-33639 346 36 nerl nerl VERB bracis-33639 346 37 ta ta X bracis-33639 346 38 account account NOUN bracis-33639 346 39 ( ( PUNCT bracis-33639 346 40 3006206169 3006206169 NUM bracis-33639 346 41 ) ) PUNCT bracis-33639 346 42 university university NOUN bracis-33639 346 43 of of ADP bracis-33639 346 44 notre notre PROPN bracis-33639 346 45 dame dame PROPN bracis-33639 346 46 hesburgh hesburgh PROPN bracis-33639 346 47 library library NOUN bracis-33639 346 48 ( ( PUNCT bracis-33639 346 49 3000184373 3000184373 NUM bracis-33639 346 50 ) ) PUNCT bracis-33639 347 1 © © ADP bracis-33639 347 2 2025 2025 NUM bracis-33639 347 3 springer springer NOUN bracis-33639 347 4 nature nature NOUN