









































Pa
ge

 
1



Pa
ge

 
12

9

American Journal of  
Geospatial Technology (AJGT)

Assessment of  the Contribution of  Geospatial Technology on Crime 
Prediction: A Case of  Kinondoni, Dar es Salaam

Nickson Alnkiza Ernest1*

Volume 4 Issue 1, Year 2025
ISSN: 2833-8006 (Online)

DOI: https://doi.org/10.54536/ajgt.v4i1.4897
https://journals.e-palli.com/home/index.php/ajgt

Article Information ABSTRACT

Received: April 01, 2025

Accepted: May 05, 2025

Published: September 22, 2025

Using quantitative methods, this article evaluates the spatial distribution of  crime, factors 
for crime occurrence and crime prediction for the Kinondoni district of  Dar es Salaam. 
The study area and participants were selected using both probability and non-probability 
sampling techniques. Participatory mapping, remote sensing, document review, GPS survey, 
and observation were used to acquire data for this study, which examined two types of  
crimes: burglaries and robberies. Hot spot maps, Kernel density maps, crime patterns, and 
visualizations were made using ArcGIS software. Areas with a high risk of  crime were iden-
tified by using kernel density, it was predicted that some areas would have high crime in the 
future, while other areas would experience low crime. It came clear that Occupation status, 
Income level and Family breakup are factors for robbery and burglary occurrence where by 
occupation status has  coefficient level of  .070, income level has coefficient level of  .067  
and coefficient level of  .878 it remains constant. The spatial distribution of  crimes in the 
Kinondon district is therefore linked to both spatial and spatial elements that cause people 
to fail in their day-to-day endeavors.

Keywords
Burglary, Crime Incidents, Hotspot 
Crime Prediction, Robbery

1 The University of  the West Indies ,Mona (Caribbean), Jamaica
* Corresponding author’s e-mail: ernestnickson1997@gmail.com

INTRODUCTION 
Various official documents and other figures based on 
police-recorded offenses revealed that homicide rates were 
high in America and that the rate of  violent, property-
related, and drug-related crimes had increased on all 
continents (UNODC, 2014). Just 5% of  the over half  
a million homicides that took place worldwide in 2012 
took place in Europe, whereas 31% took place in Africa, 
following the Americas with 36% (UNODC, 2013).
According to official and independent sources, the crime 
rate is high in African nations (UNODC, 2013). Six 
African nations were listed as having some of  the highest 
rates of  crime in the world (Numbeo, 2015). Nigeria, 
Kenya, South Africa, South Sudan, and Libya were among 
them. Robbery, corruption, consumer fraud, sexual 
assault, kidnapping, and property crimes like carjacking, 
livestock theft, and burglary were common, albeit to 
differing degrees, in other African nations like Ghana, 
Kenya, Nigeria, Egypt, Tanzania, and Uganda (UNODC, 
2014).South Africa, cases of  murder, specifically house 
robbery, and hijacking, have continued to rise in the 
country (South Sudan Monitor, 2011; Eye Witness News, 
2014; Institute for Security Studies and Africa Check, 
2014; South African Police Service, 2014).  
According to the Crime and Traffic Incidents Statistics 
Report of  January to December 2016, crime was more 
common in Tanzania’s larger cities, especially Dar es 
Salaam, with the Kinondoni district being the main focus 
of  concern. There were many different types of  crime 
in Tanzania, but the most common ones were theft, 
murder, crime against women, and the unlawful sale 

and consumption of  alcohol and drugs in public. The 
fact that cities serve as a breeding ground and favorable 
environment for organized and specialized crime is also 
no longer a secret; statistics show that Kinondoni has 
the greatest number of  recorded crimes in the nation 
(Tanzania Crime Statistics Report, 2016).Various social 
factors such as growing unemployment, increasing income 
inequality between the have and the have-nots have 
accelerated the level of  urban crimes in Kinondoni district. 
Despite their best effort the law enforcement agencies rate 
of  criminal cases is still growing fast. As a result, citizens 
live with fear insecurity of  their life and property. The rate 
of  crime has been caused by several factors which included 
political factors such as state fragility and state failure, and 
historical factors such as the history of  inter-ethnic and 
interracial injustices including apartheid. Others were 
economic factors such as unemployment and corruption, 
and the balloon effect resulting from the improved 
successes of  anti-drug law enforcement in Europe and the 
Caribbean, which are said to be responsible for increased 
drug trafficking in West Africa (UNODC, 2007; The 
Economist, 2009; Wyler & Cook, 2009; UNODC, 2015). 
Tanzania’s public and private sectors both use the geospatial 
technology system. The utilization of  location data with 
care is what makes geospatial technology indispensable 
for day-to-day tasks. For thirty years, the Tanzania Forest 
Agency (TFS) and Tanzania National Parks (TANAPA) 
have used technology to preserve endangered animals, 
such as elephants and rhinoceros (TANAPA&TFS, 2018; 
Tomkiewicz 1996; TCP 1998).
Tanzania’s 2050 national blueprint predicted that the 



Pa
ge

 
13

0

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

country’s geographic technology system would be useful 
for peace and security efforts. In order to respond to 
requests for duty service, police personnel in the field can 
easily locate themselves thanks to a global positioning 
system connected to computers. According to Okiyo 
(2021), advancements in geospatial technology have 
opened up numerous opportunities in the security sector. 
Law enforcement organizations have benefited from this 
in their attempts to combat crimes and establish public 
order. This indicates that the installation of  closed-circuit 
television surveillance cameras in different locations 
across Dar es Salaam has made a substantial contribution 
to the resolution of  security issues. Within three days, 
authorities were able to obtain a variety of  stolen items, 
according to the CCTV report. The ability of  police 
personnel in operational regions to create real-time crime 
maps facilitates the planning and implementation of  
preventative measures. It is for this reason that research 
was done on Geospatial prediction of  crime occurrence 
in urban areas Kinondoni, Dar es Salaam, Tanzania.

MATERIALS AND METHODS
 The Study Area  
Because Kinondoni District is the most populated 
district in Dar es Salaam city and has experienced 
excessive growth in size, population, and per capita 
income (Tanzania Population and Housing Census, 
2012), as well as being one of  the districts most likely to 
experience crime incidents, it was chosen as the study’s 
location. The 2014 Dar es Salaam Crime Report. Along 
with Temeke (to the far southeast) and Ilala (downtown 
Dar es Salaam), Kinondoni area is located northwest of  
Tanzania’s central business area. The Indian Ocean lies to 
the east, while Tanzania’s Pwani area lies to the north and 
west. Kinondoni is 531 km² in size. Although the Zaramo 
and Ndengereko were Kinondoni’s initial inhabitants, 
the district is now multi-tribal as a result of  increased 
urbanization. In terms of  administration, Kinondoni 
District is separated into 113 sub wards, 27 distinct wards, 
and 4 divisions.

Figure 1: Locations of  Kinondon district

Research Design And Sampling 
To gather quantitative data, the study employed a 
quantitative design. The study’s context-specificity allows 
researchers to focus on certain crime distribution areas, 
providing them with unique insights into how crime and 
geospatial technology interact in those places. Using GPS 
and additional GIS software (QGIS and ArcGIS), the 
researcher employed a quantitative technique to collect 
quantitative data about the crime scene. Spatial attributes 
were gathered. An effort was made to comprehend the 
evidence of  crime distribution cases in Kinondoni district 
using a quantitative approach; however, the ideal target 
population for this research was all of  the remanded 
individuals from Ukwamani and Mbezi beach A in police 

cells at Kawe police station. Additionally, in order to obtain 
sufficient information about the incidence of  crime, this 
study primarily targeted the population aged 25 and older 
as well as police officers. The researcher took into account 
that the 25-year-olds were part of  the working group and 
was mature enough to provide information regarding 
the prevalence of  crimes. The size of  the remanded 
individuals from the Ukwamani and Mbezi Beach police 
stations served as the basis for the sample size for this 
investigation. The study used Cochran’s method to 
determine the sample size (n) of  the remanded persons 
assuming that at least 25% of  the population committed 
crimes because there was a lot of  information available on 
the population of  remanded persons for Mbezi Beach and 



Pa
ge

 
13

1

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

Ukwamani. The following is the Cochran’s formula:
n=  z2pq/e2

Where,
e is the desired level of  precision (margin of  error) p is 
the estimated proportion of  the population  q is 1-p since 
e can take a value up to 8% n is sample size of  remanded 
persons  
A 95% confidence level gives us z-value of  1.96 per the 
normal table, so we get  
n = ((1.96)2(0.25)(0.75))/(0.08)2

≈112.55
≈ 112 Remanded persons
The Sample size of  remanded persons, n= 112, doing this 
all member of  the reminded people population had equal 
chance to be included as sample.

Data Collection 
The research made use of  both primary and secondary 
sources of  information. Remote sensing, satellites, 
and Rasta data are secondary sources of  information, 
while community mapping and field observation are 
primary sources. Pre-fieldwork and fieldwork were the 
two phases of  the study. The image, which has a 4 cm 
spatial resolution, was taken from Google Earth during 
pre-field in 2019.The basis map for the participants’ 
mapping of  crime occurrences at the scene was made 
using the photograph. The image was projected to the 
plane coordinate system, WGS_1984_Zone_36S, after 
being geometrically adjusted using the world geodetic 
system (WGS 1984). Transect mapping, observation, and 
a participatory mapping exercise were carried out during 

fieldwork. In order to demonstrate patterns of  criminal 
episodes, a Georeferenced satellite picture that was 
downloaded and covered the research region was used 
as the foundation map for participatory mapping. People 
were able to identify crime hotspots using participatory 
maps, and they also utilized them to designate regions for 
various land use activities, including businesses, schools, 
clubs, and hospitals. People were asked to identify crime 
hotspots on this participatory map of  Kawe ward, which 
was scanned in A0 format and printed off. It was then 
Georeferenced and eventually transformed into digital GIS 
software for hotspot analysis. Stakeholder identification to 
take part in the mapping exercise and the actual mapping 
session were the two steps of  the study’s methodology. 
“Participants in every police cell were identified using 
police officers during the stakeholder identification 
process. The study sample size was the basis for the 
selection criteria utilized in the questionnaire. Participants 
in the participatory mapping process included remanded 
individuals in police cells at the Kawe police station as 
well as police personnel. An experiment in participatory 
mapping was part of  the second stage. In the form of  
brief  formal inquiry guidelines, a moderator provided a 
list of  themes. Spatial data was gathered using printed 
UAV imagery. Physical locations where crimes were 
likely to occur were evacuated, and groups determined 
each point based on participants’ experiences there. The 
participants were oriented using features such as the river 
and the school buildings on the printed images. Following 
their familiarization with the map, participants use pencils 
to mark the locations of  crime episodes.

Figure 2: GPS surveying and observation

GPS surveying and observation were used in this study 
to confirm data collected through participatory mapping. 
GPS surveying and observation were utilized to confirm 
the locations of  criminal incidents and land use areas, 
including schools, clubs, and commercial districts. In 
order to confirm the validity of  the data gathered from 
key informant interviews and participatory mapping, the 
study combined observation techniques with transect 
walks, which entail GPS surveying. Both public structures 
and crime-scene locations were subjected to GPS 
surveying and observation techniques. Additionally, GPS 
was used to pinpoint the locations of  these roadways 

(Bagamoyo Road and Kawe Road), which are likewise 
regarded as significant contributors to crime incidence 
areas. A handheld GPS with an accuracy of  0–1 m was 
employed. Satellite imagery was superimposed on the 
point data gathered in order to verify the location of  
criminal episodes and analyze their spatial distribution.

Data Analysis 
In order to create maps of  crime occurrence, hotspot 
analysis was used throughout the study region to identify 
areas that were likely to experience criminal activity. Both 
spatial and social-economic aspects were evaluated. The 



Pa
ge

 
13

2

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

main purpose of  hotspot mapping analysis was to locate 
criminal incidents in the research area. The hotspots were 
designed to display trends in crime episodes, as well as 
the spatial position and amount of  criminal incidents that 
tend to focus on the research area.Using QGIS and Arc 
map software, hotspot mapping was carried out to identify 
trends in the occurrence of  infractions. Making a polygon 
grid with a cell size of  215 meters that covered the research 
region was the first step in the hotspot analysis process. Arc 
GIS was launched, and the points data layers were joined 
into the polygon grid according to spatial location. After 
that, crime hotspot locations were determined using the 
spatial statistic tool, and the analysis’s findings were shown 
so that the hotspots formed from the randomly assigned 
areas could be seen using a 99% confidence level (z-score). 
A major hotspot is indicated by a crime incidence with a 
high Z score and a small P value. A substantial cold area 
is indicated by a small P value and a low negative Z score.
The intensity of  the clustering increases or decreases with 
the Z score. No geographical grouping is implied when 
the “Z” score is zero. You can compare the “z” score with 
the range of  values for a particular confidence level to see 
if  it is statistically significant. In other words, a “z” score 
of  less than -1.96 would have been classified as statistically 
not significant at a significance level of  0.05 (95 percent 

confidence level), and a z value of  1.96 or higher is 
considered factually significant. The remaining phases 
in this investigation were completed using the Arc map 
and Qgis software. Geographic positioning systems data 
point was overlaid with wards shape file, then placed into 
an excel worked in a comma-delaminate text file inform , 
then uploaded into Arc to validate the response presented 
by participants from community mapping.

RESULTS AND DISCUSSION 
Factors Influencing Crime Occurrence 
Occupation of  Respondents   
At the 95% (P value 0.05) level of  significance, occupation 
status was found to have an impact on crime occurrence, 
with a P-value of  0.03 in the chi square representation table. 
Additionally, the results indicated that 36 respondents, or 
the bulk of  those who committed burglary crimes, were 
unemployed, 10 were employed, 7 were businessmen, 
and 0 were retired. Additionally, 41 respondents—the 
majority of  those who committed robbery crimes—were 
jobless, 10 were employed, 6 were retired, and 2 were 
businessmen. According to the findings, the majority of  
respondents who committed robbery and burglary were 
unemployed and turned to criminal activity because they 
were having a hard time making ends meet.

Table 1: Occupation status of  respondents    
Variable  Category  Crime committed P-value Chisquare

Burglary Robbery  
Occupation 
status

Unemployed  36 41

0.003 8.81 Employed  10 10 
Retired 0 6
Business men 7 2

Table 2: Income level of  respondents      
Variable  Category (Tshs) Crime committed Chisquare  P-value  

Burglary Robbery  
Income level 0 – 250000 31 44

10.12 0.04
250000 – 500000 19 7
500000- 750000 2 5
750000 – 1,000,000 1 2
Above 1050000 0 1

Income of  Respondents
According to the study’s chi square representation table, 
income level had a P-value of  0.04 and was found to 
have an impact on the incidence of  crime at the 95% (P 
value 0.05) level of  significance. Additionally, according 
to the table’s chi square representation, the majority of  
respondents (31 respondents) who committed burglary 
had low incomes from 0 to 250000 Tshs, 19 had incomes 
between 250000 and 500000, 2 had incomes between 
500000 and 750000, 1 had incomes between 750000 and 

1,000,000, and 0 had incomes above 1000000.Additionally, 
the majority of  respondents (44 respondents) who 
committed robbery had incomes between 0 and 250000 
Tshs, 7 had incomes between 250000 and 500000 Tshs, 
5 had incomes between 500000 and 750000 Tshs, 2 had 
incomes between 750000 and 1000000 Tshs, and 1 had 
incomes over 1000000 Tshs. Accordingly, the findings 
showed that both of  the respondents who committed 
robbery and burglary had low incomes and hence turned 
to illegal activity in order to survive.

Family Breakup  
The study’s chi square representation table revealed that, 

at the 95% (P value 0.05) level of  significance, family 
dissolution had a P-value of  0.04 and was found to have 



Pa
ge

 
13

3

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

an impact on the incidence of  crime. Additionally, the chi 
square representation table revealed that the majority of  
respondents (27 respondents) who committed burglary 
crimes were from married households with both parents 
present, 14 respondents were from separated or divorced 
families, and 12 respondents were from widowed families.
Additionally, the majority of  respondents (27 respondents) 
who committed robbery crimes were from bereaved 

families, followed by separated families (21 respondents) 
and married families (11 respondents). According to 
the findings, both of  the respondents who committed 
robbery and burglary were brought up in a single-parent 
household with inadequate parental supervision, which 
led to their involvement in criminal activity and deviant 
behaviors (Kubende, 2008).

Table 3: Family breakup of  respondents      
Variable  Category  Crime committed hisquare  -value  

Burglary  Robbery 
Family breakup  Married family    4 11

6.58 0.04 Separated family   27 21
Widowed family 12 27

Extent (Ranks) of  Factors for Crime Occurrence  
This section demonstrates how socioeconomic factors 
contributed to the incidence of  crime in Kawe. To 
demonstrate the magnitude of  the factors influencing 

the occurrence of  crime, the binary logistic regression 
technique was used. These factors include income, 
occupation, and family structure breakdown Level. 

Table 4: Ranks of  factors for crime occurrence     
Variables Coefficients S.E. Wald df  Pvalues Oddsratio Rank 
Occupation status .070 .215 .106 1 .02 1.072 1 
Income level .067 .242 .076 1 .02 1.069 2 
Family breakup - - - - - - - 
Constant  .878 .770 1.301 1 .254 2.406 - 

Family breakup was selected as the reference category 
from the above table. Since it seems to have a 1.072-fold 
higher likelihood of  influencing the crimes committed 
by the remanded individuals when compared to 
family breakup as a reference group, occupation status 
was selected as the most significant factor. Because 
unemployment made it difficult for people to subsist, 
more people turned to illegal activity. Additionally, the 

income level of  the remanded individuals was classified 
as the second factor in the table because, when compared 
to family separation as a reference category, it seemed 
to have a 1.069-fold higher likelihood of  influencing the 
crimes committed by the remanded individuals. 

Mapping Participatory Crime
Using geographic information technology to solve spatial 

Figure 3: Robbery crime incidents at Mbezi Beach Street in Kawe  



Pa
ge

 
13

4

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

Red points were used to indicate burglary crime events 
in figure 4. According to the map, there were burglary 
crime incidents near Tangi Bovu Street, Bondeni Street, 
Chaucha-aba, and Rungwe Street. The presence of  
unstructured houses and the Mbezi River obscured the 

view and made it difficult for a victim to flee (Taylor & 
Hochuli, 2016). Additionally, burglaries were reported near 
streets like Almas Street and Kusaga Street that were close 
to the Mwaikibaki Road. Additionally, burglary cases were 
noted in the vicinity of  the beach street.

analysis of  crime problems and other police-related 
concerns is known as “crime mapping” (Boba, 2005).
Mbezi Beach and Ukwamani Street were the two streets 
that made up Kawe ward. The opinions of  Mbezi 
respondents were collected in order to indicate locations 
on a map that correspond to crime scenes. Physical areas 
where crimes were likely to occur were evacuated, and 
responders identified all of  the places. All remanded 
individuals were asked to independently identify the 
sort of  crime committed in a particular area during this 
participatory mapping activity.

Red points were used to indicate robbery crime occurrences 
in Figure 3. According to the map, there were robbery crime 
instances near Kwa Barongo Bar and B.F.C. Park Bar. The 
crime was linked to the nearby road (Aly Syskes Road), which 
allowed thieves easy access to attack and flee, particularly at 
night (Levi, 2018). Additionally, because of  the unstructured 
houses that blocked visibility and provided no means of  
escape for a victim, robbery cases were noted along Bondeni 
Street, Tangibovu Street, Almas Street, and Rungwe Street. 
Additionally, there have been robbery cases in the beach 
region; one such example was seen on Beach Street.

Figure 4: Burglary crime incidents at Mbezi Beach A Street in Kawe    

Figure 5: Robbery crime incidents at Mbezi Beach A Street in Kawe ward      

Robbery crime occurrences were shown by red-colored 
spots in figure 5 above. According to the map, robbery 
instances happened around Bondeni Street and the “Kwa 

doctor mambo” red cross because of  the unstructured 
houses that blocked view and left victims with no way 
out. Additionally, the map showed that the presence of  



Pa
ge

 
13

5

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

Kawe Road had an impact on robbery incidents. For 
instance, robberies were seen at bus stops like Myfair 
and Ukwamani. Additionally, robberies took place near 

the well-known market area known as “Soko la Zamani” 
due to the late closing hours, which make it easier for 
criminals to target customers (Levi, 2018).

Incidents of  burglaries were shown in red spots in the above 
figure. Due to the existence of  slums and disorganized 
housing that encourage criminal activity, burglary crime 

occurrences happened near Bondeni Street and the “Kwa 
doctor mambo” Red Cross, as well as near the well-known 
“Soko la zamani” market place (Masese, 2007).

Figure 6: Showing burglary incidents at Ukwamani Street, in Kawe ward   

Table 5: Showing Summary of  Distribution of  Crime Incidents around Kawe      
Crime distribution Predominant area 

M
be

zi
 b

ea
ch

Robbery B.F.C park Kwa barongo bar along Allysyskesroad, 
MwaikibakiroadandBagamoyo road and Beach 
street

Along Ally syskes road, 
MwaikibakiroadandBagamoyo 
road 

Ro
bb

er
y

Burglary Tangibovu, Bondeni  Rungwe, Almas, Chaucha-aba 
and Beach streets

Bondeni street 

U
kw

am
an

i  
 

(K
aw

e)
 

Robbery Bondeni street, Lugalo hostel
“Kwa doctor mambo” red cross, Ukwamani bus 
stop, myfair bus stop and alongside Kawe road.  

“Kwa doctor mambo” red cross, 
bondeni street and  Ukwamani 
bus stop. 

Bu
rg

la
ry

  

Burglary Bondeni street, “Kwa doctor mambo” red cross 
and “Soko la zamani”  

“Kwa doctor mambo” red cross 
and “Soko la zamani”  

Crime Hotspots Mapping Around Kawe Ward 
Hotspot maps were divided into two streets by the study; the 
first one displayed the spatial patterns of  crime occurrences 
in Mbezi Beach Street, while the second one displayed the 
spatial patterns of  crimes in Ukwamani Street.  
Medium hotspots were displayed in yellow, while robbery 
hotspots were displayed in red. The B.F.C. Park bar, 
Barongo bar, and Beach Street were found to be the 
hotspot spots. The vicinity of  Bondeni Street, Tangibovu 
Street, Almas Street, and Rungwe Street were covered by 
medium hotspots. Based on observations made at Mbezi 
Beach A, the study found that hotspots were mostly 
found around upscale locations like beaches and bars and 
pubs that were close to the road, such as Barongo Bar and 
B.F.C. Park Bar on Ally Syskes Road.

Medium hotspots were displayed in yellow, and burglary 
hotspots in red. The regions surrounding Rungwe Street, 
Bondeni Street, and Almas Street were designated as 
hotspots, while Tangi Bovu Street, Chaucha-aba Street, parts 
of  Mwaikibaki Road, and Beach Street were recognized as 
medium hotspots. It was noted that slums with inadequate 
infrastructure bordered Rungwe and Bondeni Street, 
according to the Mbezi Beach crime hotspot map.
Hotspots were displayed in red for robbery and yellow for 
medium hotspots. Bus stops such Ukwamani bus stop, 
Bondeni street, the well-known Soko la Zamani market, 
and Kwa doctor Mambo Red Cross were identified as 
robbery hotspots. The Myfair bus station, the vicinity of  
the Ukwamani dispensary, and the Lugalo hostels were 
identified as medium hotspots.



Pa
ge

 
13

6

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

Figure 7: Robbery crime hotspot map at Mbezi Beach   

Figure 8: Burglary crime hotspot map of  Mbezi Beach A in Kawe ward   

Figure 9: Robbery crime hotspot map of  Ukwamani street Kawe ward



Pa
ge

 
13

7

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

Burglary Crime Hotspot Map of  Ukwamani Street 
in Kawe
Locations with medium hotspots were colored yellow, 
while locations with burglaries were colored red. The 

Kwa doctor Mambo Red Cross, Bondeni Street, and 
“Soko la zamani” market center have all been designated 
as burglary hotspots. Areas close to Kwa Doctor Mambo 
Red Cross were identified as medium hotspots.

Table 6: Showing summary of  hotspot distribution around Kinondoni district       
Boundary Crime type Hotspot areas Medium spot 
Mbezi 
beach  

Robbery B.F.C Park bar, Barongo bar and Beach 
street 

Bondeni street, Almas street, Rungwe street 
and Tangibovu street. 

Burglary Rungwe street, Bondeni street and 
Almas street 

Tangi bovu street, Chaucha-aba street, Beach 
street and areas along Mwaikibaki road.  

Ukwamani  Robbery  Bondeni street, Soko la zamani, 
Ukwamani bus stop and  Kwa doctor 
Mambo red cross 

Myfair bus station, areas around Ukwamani 
dispensary and Lugalo hostels 

Burglary Bondeni street, Kwa doctor Mambo 
red cross and “Soko la zamani” market 
place

Areas around to Kwa doctor mambo red 
cross.

Prediction for Crime Occurrence  
The study calculated the unknown distance of  crime 
spots in this section using interpolation techniques. One 
GIS method employed in the study to forecast cell values 
from a sample of  current data points is called spatial 
interpolation. Unknown values for any geographic point 
were ascertained by using the available data points. In 
order to forecast the incidence of  crime over the surface, 
the study used a spatial interpolation approach called 
IDW (Inverse Distance Weight). To find the absolute 

difference over the surface between current and expected 
crime patterns, the continuous surface from IDW 
interpolation was compared to the crime density map 
produced by the kernel density approach.
Additionally, the study’s crime spots were divided into 
five classes, as the table above illustrates. These classes 
represent the interpolated size of  the impact that crime 
occurrences have on victims in each area. Then, to create 
thematic maps that displayed predictable crime hotspots 
in the Kinondoni district, interpolated maps were classed 
and shown into three groups: high, medium, and low 
crime regions.
The absolute surface difference between the current and 
anticipated crime patterns at Mbezi Beach    is depicted in 
the above figure8. The current crime trends were shown 
on the Kernel Density Map (KDE), with blue denoting 
criminal risk patterns. Bondeni Street, B.F.C. Park Bar, 
Rungwe Street, and Beach Street were identified as places 
with a high crime density. Almas Street, Kusaga Street, 
Tangibovu Street, Barongo Bar, and Chaucha-aba Street 

Table 7: Showing magnitude of  crime prediction      
Class Crime Magnitude 
1 very low 
2 low 
3 medium 
4 high 
5 very high 

Figure 10: Burglary crime hotspot map of  Ukwamani street in Kawe



Pa
ge

 
13

8

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

were among the areas having a medium crime density. 
Neema Street in the vicinity of  Mbezi Beach are among 
the locations with no criminal cases. A primary school 
and the vicinity of  Mbezi Beach Rahman mosques and 
a Lutheran church. The predictive continuous smooth 
crime surface, which was divided into three classes using 
the quantile classification approach, was shown on the 
IDW interpolation map.
Bondeni Street, Tangibovu Street, B.F.C. Park Bar, 
Rungwe Street, Beach Street, and Kwa Barongo Bar were 
among the high-crime-risk zones indicated by green. 
Almas Street, Neema Street, the vicinity of  the Lutheran 
church on Mbezi Beach, Chaucha-Aba Street, and the 
region surrounding Mbezi Beach were all highlighted in 
yellow as places with medium risk of  crime. An office 
of  the local government (Mbezi Beach A MTA Office). 
Additionally, areas indicated in purple that were thought 
to have a low crime risk included the area surrounding 
Mbezi Beach. Kusaga Street, a primary school, and a 
few locations near Tangi Bovu Street and Neema Street. 

The picture above illustrates the disparity in crime 
intensity coverage surrounding Mbezi Beach between 
the Kernel density (KDE) crime map and the IDW 
interpolation crime map.The IDW interpolation map 
showed more extensive coverage of  crime patterns in 
terms of  intensity than the KDE map, such as the areas 
surrounding Rungwe Street, Bondeni Street, and Beach 
Street. Additionally, areas with no crime incidents on the 
KDE map were predicted to have high crime rates on the 
IDW interpolation map. For instance, the interpolation 
map showed low crime intensity in the vicinity of  Neema 
Street, Mbezi Beach A Primary School, and Mbezi Beach 
A Local Government Office (Mbezi Beach A mtaa 
Office).
Also areas which had low crime intensity on KDE map 
were predicted as medium crime intensity(yellow color) 
on the IDW interpolation map for example  Kusaga street 
and also location such as Tangibovu street which had 
medium crime intensity on KDE were predicted having 
high crime intensity on the IDW interpolation map. 

Figure 11: Crime prediction over surface at Mbezi Beach A in Kawe 

Figure 12: Crime prediction over surface at Ukwamani street within Kawe  



Pa
ge

 
13

9

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

The absolute surface difference between the current and 
anticipated crime trends at Ukwamani Street was depicted 
in Figure 10. The current crime trends were shown on the 
Kernel Density Map (KDE), with blue denoting criminal 
risk patterns. According to the KDE map, Bondeni Street, 
Kwa Doctor Mambo Red Cross, Soko La Zamani Market, 
and Ukwamani Bus Station were among the places with 
the highest crime densities. Areas surrounding Masjid 
Watsa are among the locations having a medium crime 
density. Lugalo Hostel and a few locations along Kawe 
Road were areas with low crime rates. Areas surrounding 
Dawasco and Lugalo are among the locations with no 
crime cases.
In addition, the IDW interpolation map showed the 
degree of  criminal risk, which was shown in green. 
Bondeni Street, Kwa Doctor Mambo Red Cross, Soko la 
Zamani Market, Ukwamani Bus Station, and the vicinity 
of  Masjid Watsa were identified as high-crime risk zones. 
Yellow-colored areas with medium-risk crime included 
the Lugalo Hostel, the roadsides, such as Bagamoyo Road 
and Kawe Road, and the vicinity of  Dawsco. Additionally, 
the purple-colored areas surrounding Lugalo Military 
Base were those that were thought to have a low crime 
risk.The crime intensity coverage surrounding Ukwamani 
Street showed a difference between the Kernel density 
(KDE) crime map and the IDW interpolation crime map. 
IDW interpolation maps showed more coverage of  crime 
patterns than KDE maps, such as those at Bondeni Street, 
Kwa Doctor Mambo Red Cross, Soko La Zamani Market, 
Ukwamani Bus Station, and the vicinity of  Masjid Watsa. 
Additionally, areas with low crime rates on the KDE 
map were predicted to have high crime rates on the IDW 
interpolation map. For instance, the areas surrounding 
Dawasco and Lugalo were predicted to have high crime 
rates on the interpolation map, while areas with medium 
crime rates on the KDE map were predicted to have high 
crime rates, such as the areas surrounding Masjid Watsa. 
Due to military security that prevents criminal activity, 
the IDW interpolation crime map forecasted low crime 
intensity around military bases (Lugalo Military Base). 
Additionally, high crime projections were discovered in 
the vicinity of  bus stations, such as Ukwamani and May 
Fair, as criminals frequently target travelers traveling by 
bus, car, or private transportation. Poor police patrol 
frequency, particularly during late hours, was cited as the 
reason for the crime prediction in the “Soko la Zamani” 
market area. Furthermore, it was predicted that crime 
would be more common near streets like Bondeni Street, 
Kwa Doctor Mambo Street, and the vicinity of  Masjid 
Watsa because these areas are primarily made up of  large 
impoverished communities living in densely populated 
squatter and unstructured settlements that make it easier 
for criminals to commit crimes. For instance, most homes 
have broken windows, roofs, and doors that let burglars 
pass through.

CONCLUSION  
Locations of  crimes and predictions were traced  on    the 

research area, which included hotspots   like the Lugalo 
Hostel, businesses, bus stations, the road zone, and 
some locations along Kawe Road. and the vicinity of  
Soko la Zamani Market, Ukwamani Bus Station, Masjid 
WatsaKwa Doctor Mambo Red Cross, Furthermore, 
it became evident that some places will experience  
high crime rates throughout time, while others will 
experience  low crime ratesas time goes on So far The 
study experienced some shortcomings and different 
conditions that could not be swayed by the researcher.  
Like Delays in issuing the approval permits from the 
Council (Kinondoni district) all led to further delays in the 
commencement of  data collection process. The problem 
was solved by compensating wasted time during field 
work, whereby the data collection process during field 
work was done on time. Unwillingness of  respondents 
especially project administrators during data collection 
was highly experienced though The problem was solved 
by ensuring the confidentiality of  the information to 
the respondents. Also, Delays and bureaucracy of  key 
respondents was a challenge in implementing the study 
due to busy schedules of  the officials during the time 
of  data collection process. The problem was solved by 
keeping the conversation with respondent shortly and 
focused on key issues to compensate the wasted time.  
From the study researcher recommends: there is a need 
for the government to recruit new intake of  polices 
and adequately trained as well as building of  new police 
stations and their residence in areas more prone to crimes 
occurrence in line with population growth, there should 
be provision of  employment to the youth in formal or 
informal sector so as they can sustain their basic needs 
this will reduce the rate of  crime occurrence in various 
spots hence the problem of  unemployment will be 
resolved as a motive towards crime occurrence.  iii. 
Furthermore, there is a need for police department to 
develop and apply software in detecting areas where there 
is high risk of  crime occurrence as well as ensuring that 
there is improving training to police officers in line with 
advancement in science and technology.  Furthermore, 
there is a need to provide civic education to the 
community so as citizen to follow placed laws and by laws 
without coercion. Furthermore, there is a need to involve 
community in security planning and in crime resolution as 
well as investigation. Moreover, provision of  equipment’s 
such as car, motorcycle and other modern investigative 
tools for the police officers is needed. And lastly, the 
government should ensure proper and normal distribution 
of  police stations and police officers in all areas with the 
introduction of  mobile stations. Hence, Further Research 
can be developed based on the presented findings since 
the study only focus on the two types of  crimes robbery 
and burglary while there are more numbers of  different 
crimes such as murder, rape, theft, child abuse, assault, 
domestic abuse, cybercrime and so more. Also, it is more 
important for other studies to base on the trend of  crime 
occurrence of  which the study was unable to identify due 
to lack of  sufficient data.  



Pa
ge

 
14

0

https://journals.e-palli.com/home/index.php/ajgt

Am. J. Geo Spat. Technol. 4(1) 129-140, 2025

REFERENCES 
Alnkiza, E. N. (2024). The role of  geospatial technology in 

Crime reduction in Dar es Salaam, Tanzania. Dar es 
Salaam: Tanzania.

Boba, R. (2005). Crime Analysis and crime mapping. Thousand 
Oaks, CA: Sage. 

Ernest, N. A., & Kyengya, A. K. (2025). Assessment of  
the Distribution of  Crimes and Factors Influencing 
Crime in the Temeke Municipality, Tanzania Using 
Geospatial Technology. American Journal of  Geospatial 
Technology, 4(1), 95–103. https://doi.org/10.54536/
ajgt.v4i1.4790 

Kyengya, A. K., Ernest, N. A., & Mtambo, C. A. (2025). 
Geospatial Technique and Forest Management of  
Coastal Areas in Bagamoyo Tanzania. American Journal 
of  Geospatial Technology, 4(1), 84–94. https://doi.
org/10.54536/ajgt.v4i1.4865

Kubende, H. (2008). Factor Influencing Crime in the 
Urban Informal Settlement: A Case of  Kibra. Journal 
on Socio Economics.

Okeyo, G. (2021). Advancing Crime Prevention, Criminal 
Justice and the Rule of  Law: Towards Achievement of  the 
2030 Agenda. Country statement of  the Republic  of  
Kenya at the fourteenth united nations congress on 
crime prevention and criminal justice. Kyoto National 
CrimeResearch Centre Nairobi; Final report 2012 .

Tanzania Crime and Safety report 2019. (2019). OSAC-
2019. CSR- Tanzania.   

Tanzania, N. (2022). Population and housing census: population 
distribution by administrative areas. Ministry of  Finance, 
Dar es Salaam.

Taylor, L., & Hochuli, D. F. ( 2016). Defining Greenspace: 
Multiple Uses Across Multiple Disciplines. Landsc. Urban 
Plan.

United Nations. (2014). World Crime Index by Country.
World Crime Report for the Period of  (2014-2015). 

United Nations Office on Drugs and Crime. (2009).
Victimization Survey in UNODC Homicide Statistics, 
(2013). Available: http://www.unodc.org/unodc/en/
data-and-Analysis/homicide.html.


