59 | International Journal of Informatics Information System and Computer Engineering 3(1) (2022) 59-70 

 

 

 

 

 

Association Analysis with Apriori Algorithm for 

Electronic Sales Decision Support System  

R. Fenny Syafariani 

 

Mathematics Department, Faculty of Ocean Technology Engineering and Informatics, Universiti 

Malaysia Terengganu, Malaysia 

 

Corresponding Email: r.fenny.syafariani@email.unikom.ac.id  

 

A B S T R A C T S  A R T I C L E   I N F O 

The purpose of this study was to determine the level of 
dependence of various items in order to dig up 
information on what items are dependent on other 
items. The method used in this research is descriptive 
analysis with a qualitative approach through a priori 
algorithm. The results show that the association 
analysis of the 26 transactions taken is 76.47%. A 
consumer who buys a laptop electronic device has the 
possibility to also buy an electronic mouse.  

 

 

 

 Article History: 
Received 25 May 2022 
Revised 30 May 2022 
Accepted 10 June 2022 
Available online 26 June 2022 
Aug 2018 

__________________ 
Keywords: 
Technology, 
Information System, 
Apriori,  
Algorithm, 
Decision Support System, 
Electronic sales 

1. INTRODUCTION 

Data mining is a data processing 
method to find patterns from the data 
obtained (Ordila et al., 2020). There are 
many methods in data mining. One 
method that is often used is the 
association method or association rule, 
more precisely using the Apriori 
Algorithm. The data generated from the 
sales process or transaction data is 

processed by association rules to find out 
information related to product purchases 
made by buyers (Riszky et al., 2019). 

There are various kinds of electronic 
goods that are sold such as Laptops, 
Printers, Mouse and so on. Sales 
transactions continue to grow every day 
and cause huge data storage (Purnia et 
al., 2017). Most sales transaction data is 
only used as an archive without being 

 

International Journal of Informatics, 
Information System and Computer 

Engineering 

International Journal of Informatics Information System and Computer Engineering 3(1) (2022) 59-70 

mailto:r.fenny.syafariani@email.unikom.ac.id


R. Fenny Association Analysis with Apriori...| 60 

 

 

 

 

used properly. However, this data set 
contains very useful information. With 
the application of association analysis or 
association rule mining in this discussion, 
it is hoped that association rules can be 
found between a combination of items. So 
that obtained a knowledge of the 
application of the concept of association 
mining analysis through the search for 
support and confidence. 

The previous research discussed 
"Application of Data Mining for Analysis 
of Consumer Purchase Patterns with the 
Fpgrowth Algorithm on Motorcycle 
Spare Part Sales Transaction Data" (Fajrin 
et al., 2018). This research is tested in 
order to influence consumer buying 
patterns, because each consumer's 
buying pattern is different. This needs to 
be analyzed further so that it can produce 
useful information, as well as maximize 
the benefits that can be obtained. Then 
the next research that has been done 
previously is discussing "Data Mining 
Analysis for Clustering Covid-19 Cases in 
Lampung Province with the K-Means 
Algorithm" (Nabila et al., 2021). This 
study was to analyze data on Covid-19 
cases in order to find out the grouping of 
the Covid-19 case problems in Lampung 
Province. The grouping of data on Covid-
19 cases in Lampung Province was 
carried out using the Clustering method 
with the K-Means algorithm. The results 
of DBI validation using manual 
calculations and using the help of 
RapidMiner tools have differences, in this 
case manual calculations have better 
results than using RapidMiner tools, but 
the results of both calculations are both 
close to 0 which means the evaluated 
clusters produce good clusters. 

In the previous research conducted 
using the fpgrowth algorithm analysis 

method, and subsequent research using 
the k-means method in conducting the 
analysis. This research is "Association 
Analysis with Apriori Algorithm for 
Decision Support System for Selling 
Electronic Goods" (Riszky et al., 2019). 
Data mining and a priori algorithms are 
very useful to find out the reality of the 
frequency of sales of electronic goods that 
are most in demand by consumers. so 
that it can be used as very useful 
information in making decisions to 
prepare stocks of what types of electronic 
goods are needed in the future.  

2. METHOD 

In this study using descriptive 
analysis method with a qualitative 
approach (Rahmawati et al., 2018). While 
in data processing using data mining 
techniques. The algorithm approach used 
is the a priori algorithm. The process of 
forming a combination of itemsets 
pattern and making rules starts from data 
analysis. The data used is data on sales of 
electronic goods, then followed by the 
formation of a combination of itemsets 
pattern and  

from an interesting combination of 
itemsets, association rules are formed. 
Then the data is made in tabular data 
format (Tana et al., 2018; Syahril et al., 
2020; Simbolon, 2019). In relation to the 
application used in the test, it is an 
application that uses one of the Microsoft 
Excel databases with data in tabular data, 
then the sales transaction data (electronic 
goods data out), is converted into binary 
form (Triansyah et al., 2018). After that 
the formation of a combination of two 
elements with a minimum value of the 
frequency of occurrence = 15 and a 
minimum value of confidence = 75%. To 



61 | International Journal of Informatics Information System and Computer Engineering 3(1) (2022) 59-70 

 

calculate support and confidence, the 
following formula is used: 

 

 

 

 

 

3. RESULTS AND DISCUSSION 

3.1. Preprocessing 

The dataset used as a test sample in 
this study uses 26 transaction data. In the 
data there are several items of electronic 
goods sold, namely printers, laptops, 
chargers, and mice. The following is a 
table of transaction data that is used as a 
sample.  

Table 1. Transaction Data Table 

Transaction Item 

E1 Printer,Laptop 

E2 Laptop, Printer 

E3 Charger, Printer 

E4 Printer, mouse, Printer 

E5 Charger, Printer,Printer 

E6 Laptop, Printer 

E7 Printer,Laptop,Charger 

E8 Printer, Laptop, Printer 

E9 Printer,Laptop 

E10 Printer, Printer 

E11 Printer, Laptop, Printer 

E12 Laptop, Printer 

E13 Charger, Printer,Printer 

E14 Printer,Laptop 

E15 Printer,Laptop,Charger 

E16 Printer, Charger 

E17 Charger, Printer 

E18 Charger, Printer,Printer 

E19 Printer, Laptop, Printer 

Support = 
𝛴𝑖𝑡𝑒𝑚𝑠 𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑 𝑎𝑡 𝑜𝑛𝑐𝑒

𝛴𝑡𝑜𝑡𝑎𝑙 𝑡𝑟𝑎𝑛𝑠𝑎𝑐𝑡𝑖𝑜𝑛
× 100% 

Confidence  = 
𝛴𝑖𝑡𝑒𝑚𝑠 𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑 𝑎𝑡 𝑜𝑛𝑐𝑒

𝛴𝑡𝑟𝑎𝑛𝑠𝑎𝑐𝑡𝑖𝑜𝑛𝑠 𝑖𝑛 𝑡ℎ𝑒 𝑎𝑛𝑡𝑒𝑐𝑒𝑑𝑒𝑛𝑡 𝑠𝑒𝑐𝑡𝑖𝑜𝑛
× 100 % 



R. Fenny Association Analysis with Apriori...| 62 

 

 

 

 

E20 Charger, Printer,Printer 

E21 Laptop, Printer 

E22 Printer,Laptop,Charger 

E23 Printer,Laptop 

E24 Printer, mouse, Printer 

E25 Laptop, Printer 

E26 Printer, mouse, Printer 

 

3.2. Transaction Data Tabular Format 

The application used in the test is a 
Microsoft Excel database so that the data 
must be converted into binary form 
(Sianturi et al., 2018).The conversion 
process is that the slip number of the data 
to be tested is made horizontally 
downwards, while all types of items will 
become vertical attributes, so that they 

form like a table, based on real sales data 
(electronic goods data out) the meeting 
point between the name of the electronic 
type and the number slip will become 
binary 1, while those that do not become 
meeting points will become binary 0. The 
results of the conversion process of sales 
transaction data to data format in tabular 
data form are as shown in the following 
table:  

 

Table 2. Data in the form of tabular data 

Transaction Printer Mouse Laptop Charger 

E1 1 0 1 0 

E2 0 1 1 0 

E3 1 0 0 1 

E4 1 1 1 0 

E5 1 1 0 1 

E6 0 1 1 0 

E7 0 1 1 1 

E8 1 1 1 0 

E9 1 0 1 0 

E10 1 1 0 0 



63 | International Journal of Informatics Information System and Computer Engineering 3(1) (2022) 59-70 

 

E11 1 1 1 0 

E12 0 1 1 0 

E13 1 1 0 1 

E14 1 0 1 0 

E15 0 1 1 1 

E16 0 1 0 1 

E17 1 0 0 1 

E18 1 1 0 1 

E19 1 1 0 0 

E20 1 1 0 1 

E21 0 1 1 0 

E22 0 1 1 1 

E23 1 0 1 0 

E24 1 1 1 0 

E25 0 1 1 0 

E26 1 1 1 0 
 

17 20 17 10 

 

In the tabular data table above, the 
number of occurrences (electronic items 
that come out) of each item is: Printer = 
17, Mouse = 20, Laptop = 17, and Charger 
= 10.  

3.3.  Formation of Two Elements 
Combination Pattern 

With a minimum value of the frequency 
of occurrence Ф= 15. In the form table 

tabular data, there is one electronic item 
that does not meet the provisions of the 
frequency limit, namely Charger = 10, so 
in the formation of the pattern of these 
two elements we make a combination of 
pairs of 3 electronic items, namely 
Printer-Mouse, Laptop Printer, Mouse-
Laptop. The following are tables of 2 
element combinations:

 

Table 3. Two Elements Combination Pattern (Printer, Mouse) 

Transaction Printer Mouse f 



R. Fenny Association Analysis with Apriori...| 64 

 

 

 

 

E1 1 0 S 

E2 0 1 S 

E3 1 0 S 

E4 1 1 P 

E5 1 1 P 

E6 0 1 S 

E7 0 1 S 

E8 1 1 P 

E9 1 0 S 

E10 1 1 P 

E11 1 1 P 

E12 0 1 S 

E13 1 1 P 

E14 1 0 S 

E15 0 1 S 

E16 0 1 S 

E17 1 0 S 

E18 1 1 P 

E19 1 1 P 

E20 1 1 P 

E21 0 1 S 

E22 0 1 S 

E23 1 0 S 

E24 1 1 P 

E25 0 1 S 

E26 1 1 P 

Total (P)   11 

Table 4. Two Elements Combination Pattern (Printer, Laptop)  



65 | International Journal of Informatics Information System and Computer Engineering 3(1) (2022) 59-70 

 

Transaction Printer Laptop f 

E1 1 1 P 

E2 0 1 S 

E3 1 0 S 

E4 1 1 P 

E5 1 0 S 

E6 0 1 S 

E7 0 1 S 

E8 1 1 P 

E9 1 1 P 

E10 1 0 S 

E11 1 1 P 

E12 0 1 S 

E13 1 0 S 

E14 1 1 P 

E15 0 1 S 

E16 0 0 S 

E17 1 0 S 

E18 1 0 S 

E19 1 0 S 

E20 1 0 S 

E21 0 1 S 

E22 0 1 S 

E23 1 1 P 

E24 1 1 P 

E25 0 1 S 

E26 1 1 P 

Total (P)   9 

 

Table 5. Two Elements Combination Pattern (Printer, Laptop)  

Transaction Mouse Laptop f 

E1 0 1 S 

E2 1 1 P 

E3 0 0 S 

E4 1 1 P 

E5 1 0 S 

E6 1 1 P 

E7 1 1 P 

E8 1 1 P 

E9 0 1 S 

E10 1 0 S 



R. Fenny Association Analysis with Apriori...| 66 

 

 

 

 

E11 1 1 P 

E12 1 1 P 

E13 1 0 S 

E14 0 1 S 

E15 1 1 P 

E16 1 0 S 

E17 0 0 S 

E18 1 0 S 

E19 1 0 S 

E20 1 0 S 

E21 1 1 P 

E22 1 1 P 

E23 0 1 S 

E24 1 1 P 

E25 1 1 P 

E26 1 1 P 

Total (P)   13 

 

From the tables of the 2 elements above, 
P means that the items are sold together, 
while S means that there are no items that 
are sold together or there is no 
transaction. Σ represents the number of 
Frequency items set. So that in the pattern 
of these two elements, the support value 
is obtained, namely:  

• Printer – Mouse = 11 

• Printer – Laptop = 9 

• Mouse – Laptop = 13 
 3.4. Formation of Three Elements 
Combination Pattern 

 The combination of the 2 elements in the 
table above, we can combine into 3 
elements. For the set formed on these 3 
elements are Laptop, Printer, Mouse. 
Here is a table of 3 elements:

  

Table 6. Three Elements Combination Pattern (Printer, Mouse, Laptop) 

Transaction Printer Mouse Laptop f 

E1 1 0 1 S 

E2 0 1 1 S 

E3 1 0 0 S 

E4 1 1 1 P 

E5 1 1 0 S 



67 | International Journal of Informatics Information System and Computer Engineering 3(1) (2022) 59-70 

 

E6 0 1 1 S 

E7 0 1 1 S 

E8 1 1 1 P 

E9 1 0 1 S 

E10 1 1 0 S 

E11 1 1 1 P 

E12 0 1 1 S 

E13 1 1 0 S 

E14 1 0 1 S 

E15 0 1 1 S 

E16 0 1 0 S 

E17 1 0 0 S 

E18 1 1 0 S 

E19 1 1 0 S 

E20 1 1 0 S 

E21 0 1 1 S 

E22 0 1 1 S 

E23 1 0 1 S 

E24 1 1 1 P 

E25 0 1 1 S 

E26 1 1 1 P 

Total (P)       5 

 

It can be seen from the pattern table of the 
3 elements above, the items that were 
sold simultaneously were Laptop – 
Printer - Mouse = 5 

So, the support value in the 3 element 
pattern table is 5.  

3.4. Association Rules 

Calculating the support and confidence 
values of each frequent itemset so that 
candidate association rules appear 
(Lestari, 2017). To calculate support and 
confidence, the following formula is 
used:  

 Support = 
𝛴𝑖𝑡𝑒𝑚𝑠 𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑 𝑎𝑡 𝑜𝑛𝑐𝑒

𝛴𝑡𝑜𝑡𝑎𝑙 𝑡𝑟𝑎𝑛𝑠𝑎𝑐𝑡𝑖𝑜𝑛
× 100% 



R. Fenny Association Analysis with Apriori...| 68 

 

 

 

 

So that the results are obtained as in the 
table below 

Table 7. Association Rules Candidate List 

If antecedent, then consequent Support  Confidence 

Printer, Mouse 11/26*100 %= 42,30% 11/17*100 %= 64,70% 

Mouse, Printer 11/26*100 %= 42,30% 11/20*100 %= 55% 

      

Printer, Laptop 9/26*100%=34,61% 9/17*100%=52,94% 

Laptop, Printer 9/26*100%=34,61% 9/17*100%=34,61% 

      

Mouse, Laptop 13/26*100%=50% 13/20*100%=65% 

Laptop, Mouse 13/26*100%=50% 13/17*100%=76,47% 

 

From the table above, the support and 
confidence have been determined. then 
select the association rules that meet the 

minimum confidence of 75%, so that the 
association rules are obtained, which are 
as follows:

Table 8. Association Rules List 

If antecedent, then consequent Support  Confidence 

Laptop, Mouse 13/26*100%=50% 13/17*100%=76,47% 

From the results of the analysis that 
has been carried out, there is 1 product 
association rule that meets the minimum 
confidence limit, namely Laptop - Mouse. 
Then the results obtained are "76,47% of 
transactions that contain Laptop 
electronics also contain Mouse 
electronics. And 50% of all transactions 
that contain these two items". With 
Apriori Algorithm analysis can be 
applied to assist marketing strategies in a 
company or institutions. Data mining 
and a priori algorithms are very useful to 
find out the relationship between the 
frequency of sales of electronic goods that 
are most in demand by customers, so that 
they can be used as very valuable 
information in making decisions to 

prepare stocks of what types of electronic 
goods are needed in the future. 

4. CONCLUSION 

A priori algorithm is used in 
conducting association analysis to 
determine the level of dependence of 
various items to explore information on 
what items have dependence on other 
items based on 26 transaction records that 
are sampled. The author performs an 
association analysis calculation from the 
samples taken so that the result is that 
76.47% of a consumer who buys laptop 
electronics has the possibility to also buy 
mouse electronics. And 50% of all 
transactions that contain these two items.  

Confidence = 
𝛴𝑖𝑡𝑒𝑚𝑠 𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑 𝑎𝑡 𝑜𝑛𝑐𝑒

𝛴𝑡ℎ𝑒 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑟𝑎𝑛𝑠𝑎𝑐𝑡𝑖𝑜𝑛𝑠 𝑖𝑛 𝑡ℎ𝑒 𝑎𝑛𝑡𝑒𝑐𝑒𝑑𝑒𝑛𝑡 
𝑠𝑒𝑐𝑡𝑖𝑜𝑛

× 100 % 



69 | International Journal of Informatics Information System and Computer Engineering 3(1) (2022) 59-70 

 

REFERENCES 

Ordila, R., Wahyuni, R., Irawan, Y., & Sari, M. Y. (2020). Penerapan Data Mining 

Untuk Pengelompokan Data Rekam Medis Pasien Berdasarkan Jenis Penyakit 

Dengan Algoritma Clustering (Studi Kasus: Poli Klinik Pt. Inecda). Jurnal Ilmu 

Komputer, 9(2), 148-153. 

Riszky, A. R., & Sadikin, M. (2019). Data Mining Menggunakan Algoritma Apriori 

untuk Rekomendasi Produk bagi Pelanggan. Jurnal Teknologi dan Sistem 

Komputer, 7(3), 103-108. 

Purnia, D. S., & Warnilah, A. I. (2017). Implementasi Data Mining Pada Penjualan 

Kacamata Menggunakan Algoritma Apriori. IJCIT (Indonesian Journal on 

Computer and Information Technology), 2(2). 

Rahmawati, F., & Merlina, N. (2018). Metode Data Mining Terhadap Data Penjualan 

Sparepart Mesin Fotocopy Menggunakan Algoritma Apriori. PIKSEL: 

Penelitian Ilmu Komputer Sistem Embedded and Logic, 6(1), 9-20. 

Triyansyah, D., & Fitrianah, D. (2018). Analisis Data Mining Menggunakan Algoritma 

K-Means Clustering Untuk Menentukan Strategi Marketing. InComTech: Jurnal 

Telekomunikasi dan Komputer, 8(3), 163-182. 

Nabila, Z., Isnain, A. R., Permata, P., & Abidin, Z. (2021). Analisis Data Mining Untuk 

Clustering Kasus Covid-19 Di Provinsi Lampung Dengan Algoritma K-

Means. Jurnal Teknologi Dan Sistem Informasi, 2(2), 100-108. 

Riszky, A. R., & Sadikin, M. (2019). Data Mining Menggunakan Algoritma Apriori 

untuk Rekomendasi Produk bagi Pelanggan. Jurnal Teknologi dan Sistem 

Komputer, 7(3), 103-108. 

Sianturi, F. A. (2018). Penerapan Algoritma Apriori Untuk Penentuan Tingkat 

Pesanan. Jurnal Mantik Penusa, 2(1). 

Lestari, N. (2017). Penerapan Data Mining Algoritma Apriori Dalam Sistem Informasi 

Penjualan. Jurnal Edik Informatika Penelitian Bidang Komputer Sains dan 

Pendidikan Informatika, 3(2), 103-114. 

Tana, M. P., Marisa, F., & Wijaya, I. D. (2018). Penerapan Metode Data Mining Market 

Basket Analysis Terhadap Data Penjualan Produk Pada Toko Oase 

Menggunakan Algoritma Apriori. JIMP-Jurnal Informatika Merdeka 

Pasuruan, 3(2). 

Syahril, M., Erwansyah, K., & Yetri, M. (2020). Penerapan Data Mining untuk 

menentukan pola penjualan peralatan sekolah pada brand wigglo dengan 



R. Fenny Association Analysis with Apriori...| 70 

 

 

 

 

menggunakan algoritma apriori. Jurnal Teknologi Sistem Informasi Dan Sistem 

Komputer TGD, 3(1), 118-136. 

Simbolon, P. H. (2019). Implementasi Data Mining Pada Sistem Persediaan Barang 

Menggunakan Algoritma Apriori (Studi Kasus: Srikandi Cash Credit 

Elektronic dan Furniture). JURIKOM (Jurnal Riset Komputer), 6(4), 401-406. 

Fajrin, A. A., & Maulana, A. (2018). Penerapan Data Mining Untuk Analisis Pola 

Pembelian Konsumen Dengan Algoritma Fp-Growth Pada Data Transaksi 

Penjualan Spare Part Motor. Kumpulan Jurnal Ilmu Komputer (KLIK), 5(01), 1-10.