Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti Classification of Types of Crimes Against Human Physique Using the KMeans Clustering Method Rizkah Fadillah1,*.
Fiqri Hidayat Rangkuti2.
Anggi Jaya Maulana Siregar2.
Rizky Ananda2.
Muhammad Syahrizal3,* Program Studi Teknik Informatika.
Fakultas Ilmu Komputer dan Teknologi Informasi.
Universitas Budi Darma.
Medan.
Indonesia Program Studi Teknologi Informasi.
Fakultas Ilmu Komputer dan Teknologi Informasi.
Universitas Budi Darma.
Medan.
Indonesia Prodi Teknologi Rekayasa Komputer Grafis.
Politeknik Cendana.
Medan.
Indonesia Jl.
Sisingamangaraja No.
Siti Rejo I.
Kec.
Medan Kota.
Kota Medan.
Sumatera Utara.
Indonesia Jl.
Williem Iskandar No.
Sidorejo Hilir.
Kec.
Medan Tembung.
Kota Medan.
Sumatera Utara.
Indonesia Email: 1rizkahfadillah52@gmail.
com 2panditorangkuti@email.
com, 3Anggijaya003@email.
com, 4dhitisyhla@email.
Email Penulis Korespondensi: 5,*m.
syahrizal156@gmail.
AbstractOeHuman physical crimes are unlawful acts and prohibited by the rule of law, which can harm or damage the body of others.
This study aims to examine the number of groupings of types of crimes against human physique in 2019 to 2020 in all areas of East Nusa Tenggara province.
To do this, we use the K-Means Clustering method to group the types of physical crimes against humans.
The data used came from the Central Statistics Agency of East Nusa Tenggara province.
The K-Means method is one of the nonhierarchical data clustering methods that seeks to partition data into the form of one or more clusters/groups.
After the application of the K-Means algorithm in the grouping of types of crimes against human bodies in 2019 to 2020 in the East Nusa Tenggara region, there are 3 centroids.
C1 for areas with low crimes.
C2 for areas with moderate crimes and C3 for areas with high crimes.
The initial centroid value is determined randomly and then for the next centroid is adjusted to the result of the calculation of the closest distance .
The final results obtained are areas with low crime totaling 13 regions, namely East Sumba.
Lembata.
Sikka.
Ende.
Ngada.
Manggarai.
Rote Ndao.
West Manggarai.
Central Sumba.
Southwest Sumba.
Nagekeo.
East Manggarai, and Sabu Raijua.
There are 7 areas with moderate crimes, namely West Sumba.
Kupang.
South Central Timor.
North Central Timor.
Belu.
Alor, and East Flores.
As for the area with high crime, there is 1 area, namely Kupang City.
Keywords: East Nusa Tenggara.
Crime.
Data Mining.
Clustering.
K-Means.
INTRODUCTION
This study aims to examine the number of groupings of types of crimes against human physique in 2019 to 2020 in all areas of East Nusa Tenggara province.
Human physical crimes are unlawful acts and are prohibited by the rule of law, which can harm or damage the body of others.
The definition of a human physical crime can be known through the term "strafbaar feit" which comes from the Dutch language, which means "an unlawful act committed intentionally or unintentionally by a person whose actions can be held accountable and which are made by a person who can be held Crimes against human bodies in East Nusa Tenggara province are complex problems and require serious Various forms of violence, such as violence against children, human trafficking, and other crimes, have become part of people's lives in the province.
The factors that cause physical violence against children in Kupang City, such as lack of legal protection and lack of community participation, have become a major concern in research and This research will help those responsible for managing human physical crimes in the province of East Nusa Tenggara.
In addition, this research will also help the responsible parties in developing strategies for the prevention and control of crimes against humans in East Nusa Tenggara province.
By knowing the number of groupings of types of physical crimes against humans, we can improve the system of error management and prevention of physical crimes against humans in the province of East Nusa Tenggara.
In this study, we will conduct statistical analysis and data collection from the Central Statistics Agency of East Nusa Tenggara province.
We will also conduct a comparison and comparison of data from year to year.
By conducting an in-depth analysis, we will find trends and similarities in the number of types of crimes against physical humans in the province of East Nusa Tenggara.
To do this, a technique called data mining is needed.
Data mining is a process that uses machine learning, statistical engineering, mathematics, and artificial intelligence to identify valuable information from various large databases.
Data mining is a systematic process of extracting useful information from a set of data consisting of unknown knowledge by manual means.
Data mining allows us to identify hidden patterns in data, so that it can provide deeper insight and understanding of a phenomenon or problem.
In simple terms, data mining can be said to be the process of filtering or "mining" knowledge from a large amount of data.
Another term for data mining is Knowledge Discovery in Database (KDD).
Although data mining itself is part of the KDD process stages.
Data mining is an analytical process designed to explore large amounts of data for valuable, consistent, and hidden knowledge.
Data mining is a series of processes that extract previously unknown knowledge from a set of data.
Data mining is used to improve previous techniques so that it can handle various problems that are often encountered.
The methods used in data mining include several main classifications, namely Classification.
Clustering.
Association.
Regression, and Forecasting.
Based on several groupings in data mining, this research uses the K-Means clustering method in classifying crimes that occur in NTT.
There are several previous studies or previous studies that have been carried out so that they can be a reference in completing this research.
The research conducted in 2020 by Wahyu Saputro, and several other research colleagues.
Rizkah Fadillah.
JurIKTI | Page 1 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti where they conducted research on corruption crimes by utilizing the K-Means algorithm for classification, they conducted research in jurisdictions in Indonesia.
The final result obtained from the research they have conducted is that they formed 3 clusters which show that the DBI value obtained is good, which is 0.
113 so that it can be concluded that as many as 5 areas are prone to corruption on the island of Samatera, 3 areas are prone to corruption on the island of Java and Sulawesi and one area is prone to corruption on the island of Kalimantan.
The research conducted in 2023 by Ikhlasul Amal and Raissa Amanda Putri, where they conducted research on drug addicts using the K-Means algorithm for classification, they conducted research in the North Sumatra region, precisely in the city of Medan.
The final result obtained from the study is that they formed 4 clusters, so it can be concluded that there are 2 sub-districts that are included in the area with the highest level of drug addicts, 7 sub-districts that are included in the area with a high level of addicts, 7 sub-districts that are included in the area with a low level of drug addicts, and 5 sub-districts that are included in the area with a very low level of addicts.
Based on research conducted in 2022 by Nisriina Nuur Hasanah and Agus Sidiq Purnomo, they conducted research on the implementation of data mining for library book grouping using the K-Means algorithm method, the research was conducted at the LPP Yogyakarta Polytechnic Library.
The final result obtained from the study is to form 3 clusters, so that the results obtained are 3 books that are most in demand, 9 books that are quite in demand, and 18 books that are in little interest.
In 2021, a study was conducted by Lalu Ganda Rady Putra and Anthony Anggrawan, they conducted a study on social assistance recipients who deserve to receive assistance and those who are less eligible to receive assistance using the K-Means algorithm method for grouping, the research was conducted in Ampenan district.
West Nusa Tenggara The final result obtained from the study is that they form 2 clusters, so it can be concluded that out of 257 data, there are 196 data with the status of social assistance recipients on target and 61 data with the status of social assistance recipients who are not on target.
RESEARCH METHODOLOGY
1 Research Stages At this stage, it explains the overall description of the research method, the steps of this stage can be seen in figure 1.
Figure 1.
Research Stages Study book Literature study is a method of collecting information and data with the help of various kinds of literature in the library such as documents, books, and research journals.
Data collection At this stage, data was collected from human physical crime data in NTT in 2019-2020 obtained from the Central Statistics Agency (BPS) and the National Crime Information Center (Pusikna.
of the National Police Criminal Investigation Department.
Data Preprocessing and Data Transformation At this stage, value transformation or weighting of the original data is carried out, weighting the data from nominal data to numerical data to make it easier to calculate the K-Means algorithm.
Application of the K-Means Method Rizkah Fadillah.
JurIKTI | Page 2 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti Applying the K-Means algorithm to classify the types of crimes against human physicality.
With the aim of increasing efficiency and accuracy in the analysis of data on physical crimes against humans.
Visualization The next stage of this research is the visualization of data that has been weighted by the application of the K-Means algorithm method.
The data visualized is data on the Types of Crimes Against Human Physique in 2019 and 2020.
Analyze the results In this study, the last stage is the analysis of results.
The analysis of the results was carried out to determine the success and conclusions of the visualization that had been implemented.
2 Clustering Clustering is a process flow where a set of information subjects is put into a set of sections called clusters.
Icons or subjects in clusters have similar characters to each other and are different from other clusters.
Clustering is a technique that is often used to group data into separate groups.
The main goal is to separate and group data that have similarities or similarities in nature or characteristics among the data present in the dataset.
The clustering process is used to divide data into classes or clusters based on their degree of similarity.
Clustering in data mining is grouping a set of data or objects into clusters .
so that all data in a cluster contains data that is as similar as possible to objects in other clusters.
Two methods of grouping are known: hierarchical grouping and partitioning.
Clustering is also defined as the process of grouping the same data into different groups, or rather partitions of a set into subsets, so that the data in each subset has a useful meaning.
The way clustering works has several processes of grouping data into several clusters, so that the data in the cluster has the greatest similarity which also makes it possible to determine and retrieve data between different clusters that show minimal similarities, can also be used to determine in identifying cluster data groups that result from grouping small items based on their similarities.
3 Algoritma K-Means The k-means method is a clustering method where the cluster center uses the average value of the data in each cluster K-Means is a non-hierarchical data clustering method that seeks to partition existing data into one or more clusters/groups.
This method partitions data into clusters or groups so that data that have the same characteristics are grouped into the same cluster.
K-Means is a data analysis technique or known as the data mining technique to apply the unsupervised data modeling process and is one of the techniques to group each data in the form of partitions.
In the KMeans method, the data is grouped into several groups where each group has similar or same characteristics as the others but with other groups have different characteristics.
K-Means is a non-hierarchical method, in the K-Means process stage the cluster center is randomly selected from a set of component data in the data population and marks the component to one of the cluster centers that has been defined depending on the minimum distance between the components and each cluster.
Some of the advantages of the K-Means algorithm include being easy to implement, having a high level of convergence, and producing denser clusters when compared to hierarchical methods.
The calculation process in the K-Means algorithm has a flowchart flow shown in Figure 2.
Figure 2.
K-Means Algorithm Flowchart Rizkah Fadillah.
JurIKTI | Page 3 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti The following are the research steps carried out for sampling so that the data used can be processed, including:
Specify K as the number of clusters to be formed Determine the initial centroid.
Formula of the equation .
yaycn = " Oc" #$! ycU# .
Calculate the distance of each input data for each centroid using the Euclidean Distance formula until the closest distance from each data to the centroid is found.
Formula equation .
cu, y.
= ,Oc&'$!.
cu! Oe yc! )% .
Information:
D = Distance X= Data Y= Centroid Classify data based on their proximity to centroids.
Recount the cluster center with the current cluster members.
The cluster center is the average value of all the object data in a given cluster.
Count each object again using the new cluster center.
If the cluster center does not change anymore then the clustering process is complete.
Alternatively, go back to step number 3 until the center of the cluster does not change again.
RESULTS AND DISCUSSION
1 Data collection In this study, we collected data from the Central Statistics Agency of East Nusa Tenggara province.
The data we used in this study is the number of groupings of types of crimes against human physicality, in 2019-2020.
Table 1.
Number of Classifications of Types of Crimes Against Human Physicality, 2019-2020
Region
West Sumba
East Sumba
Kupang
South Central Timor
North Central Timor
Belu
Alor
Lembata
East Flores Sikka
End
Ngada
Manggarai Rote Ndao
West Manggarai Central Sumba
Southwest Sumba
Nagekeo
East Manggarai Sabu Raijua
Kupang City East Nusa Tenggara
Number of Classifications of Types of Crimes Against Human Physique Murder
Rape
Normal/Mild Persecution Severe Persecution Abduction
2019 2020 2019 2020
2 Application of K-Means Clustering Method Iteration 1 Number of clusters C=3 (C1.
C2, and C.
Early cluster centroid The initial centroid center was used for the grouping or cluster used in this study.
The initial centroid center is determined randomly, so the data used is free to any data.
For the early centroid center, it can be seen in the following table 2.
Rizkah Fadillah.
JurIKTI | Page 4 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti Table 2.
Early Centroid Center(Early Centroi.
Region
East Sumba
Alor
Kupang City Number of Classifications of Types of Crimes Against Human Physique Murder
Rape
Normal/Mild Persecution Severe Persecution Abduction
2019 2020 2019 2020
yaycn = " Oc" #$! ycU# Calculate the distance of each centroid's input data using the Euclidean Distance formula ya.
cu, y.
= ,Oc&'$!.
cu! Oe yc! )% Data 1:
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya1 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 112,4188596 .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya2 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 98,32598843 .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya3 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 381,5193311 Data 2:
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya1 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 0 .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya2 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 209,5853048 .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya3 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 491,3318634 Data 3:
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya1 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 170,449406 .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya2 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 55,88380803 .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% ya3 = / .
Oe .
% .
Oe .
% .
Oe .
% .
Oe .
% = 321,7452408 Perform the above calculation steps until the 21st data.
The following is the closest distance based on the initial centroid can be seen in the following table 3.
Table 3.
Minimum Distance of Iteration 1 Region West Sumba East Sumba Kupang South Central Timor North Central Timor Belu Alor Lembata East Flores 112,4188596 170,449406 154,5153714 138,6145735 108,9632966 209,5853048 39,61060464 104,885652 98,32598843 209,5853048 55,88380803 64,86139067 88,81441324 123,5839795 233,6985237 105,5604092 381,5193311 491,3318634 321,7452408 337,8579583 382,3519321 385,8134264 294,4401467 521,4019563 393,9390816 Nearby 98,32598843 55,88380803 64,86139067 88,81441324 108,9632966 39,61060464 104,885652 Cluster Rizkah Fadillah.
JurIKTI | Page 5 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti Region
Sikka
End
Ngada
Manggarai Rote Ndao
West Manggarai Central Sumba
Southwest Sumba
Nagekeo
East Manggarai Sabu Raijua
Kupang City 28,53068524 41,36423576 37,82856064 51,623638 32,54228019 33,91164992 102,5962962 59,78294071 52,62128847 49,49747468 61,16371473 491,3318634 229,7476877 186,4215653 177,0056496 159,0188668 236,4931289 234,4397577 183,2484652 263,5602398 253,1264506 243,2036184 267,2395929 294,4401467 515,4716287 475,9936974 464,1379105 440,9920634 521,6895629 520,6755228
476,8658092
548,6647428
539,9833331
531,602295
551,8532414
Nearby 28,53068524 41,36423576 37,82856064 51,623638 32,54228019 33,91164992 102,5962962 59,78294071 52,62128847 49,49747468 61,16371473 Cluster The data has been classified and grouped into centroids based on the closest distance .
inimum distanc.
of the clusters that have been calculated.
After clustering clusters based on the closest distance, then perform iteration calculations using the new centroid The new centroid value is adjusted to the location of the nearest cluster and then do the equation one.
yaycn = " Oc" #$! ycU# Murder:
4 0 1 2 1 0 1 0 0 0 0 0 0 .
= 0,733333333
3 2 3 1 1 0 3 0 0 3 8 1 1 .
= 1,933333333
= Rape:
= .
1 0 3 1 1 4 0 3 0 0 0 0 0 .
= 0,933333333 .
4 0 1 5 2 1 0 6 7 7 7 0 2 .
= 3,2 ya1.
= ya1.
= Normal/Mild Persecution:
143 10 99 21 36 53 81 19 16 0 0 0 0 .
= 35,13333333
98 42 135 38 82 80 83 31 36 133 12 25 38 .
= = 58,86666667 ya1.
= Severe Persecution:
0 2 2 0 0 1 0 0 0 0 0 0 0 .
= 0,333333333
1 0 0 1 1 1 0 1 0 4 13 0 0 .
= 1,466666667
= ya1.
= Abduction:
0 0 0 0 0 0 0 0 0 0 0 0 0 .
= 0,066666667
0 0 0 0 0 2 2 0 0 2 1 0 0 .
= 0,533333333
= ya1.
= Perform the above steps to find the C2 and C3 centroids.
The following is a table of 4 initial Centroids that have been calculated based on equation .
Table 4.
Centroid Early Iteration 2
Murder
0,7333333
1,93333333
Number of Classifications of Types of Crimes Against Human Physique Normal/Mild Rape
Severe Persecution Persecution 0,93333333 35,13333 58,8666666 0,333333 1,466666 Abduction
0,06666666
0,53333333
Rizkah Fadillah.
JurIKTI | Page 6 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti Murder Number of Classifications of Types of Crimes Against Human Physique Normal/Mild Rape Severe Persecution Persecution Abduction Recalculate each object using the new cluster center.
If the cluster center does not change anymore then the clustering process is complete.
Alternatively, go back to step number 3 until the center of the cluster does not change again.
After the calculation is carried out, the iteration calculation process stops at the calculation of the 5th iteration, where the results of the 4th and 5th iterations are the same .
he location of the centroid or the nearest distance does not change or is fixe.
The following can be seen in table 5, the closest distance of the 4th iteration and the closest distance of the 5th iteration can be seen in table 6.
Table 5.
Closest Distance of Iteration 4
Region
West Sumba
East Sumba
Kupang
South Central Timor
North Central Timor
Belu
Alor
Lembata
East Flores Sikka
End
Ngada
Manggarai Rote Ndao
West Manggarai Central Sumba
Southwest Sumba
Nagekeo
East Manggarai Sabu Raijua
Kupang City 131,2380682 32,4771287 191,5222282 175,1583014 146,0688099 135,8943115 225,0639107 9,836863773 120,1829046 10,36487123 39,65808731 48,90395917 73,45132553 16,80567034 12,09396084 87,54292598 41,36541094 28,31249234 19,5387791 43,8930202 512,612359 17,31193591 128,4112656 46,45646484 30,52381243 52,24416833 50,94068242 82,35565023 155,6388548 29,12821871
150,4109475
110,2381655
98,30286428
77,35924718
156,9122466
155,4411565
126,5699535
184,1988141
174,5349625
165,7670146
187,451202
366,1109164
381,5193311
491,3318634
321,7452408
337,8579583
382,3519321
385,8134264
294,4401467
521,4019563
393,9390816
515,4716287
475,9936974
464,1379105
440,9920634
521,6895629
520,6755228
476,8658092
548,6647428
539,9833331
531,602295
551,8532414
Nearby 17,31193591 32,4771287 46,45646484 30,52381243 52,24416833 50,94068242 82,35565023 9,836863773 29,12821871 10,36487123 39,65808731 48,90395917 73,45132553 16,80567034 12,09396084 87,54292598 41,36541094 28,31249234 19,5387791 43,8930202 Cluster Nearby 27,77974596 28,20539627 37,37264087 22,48491558 53,21116961 57,93962375 71,52621897 14,96939679 38,87893297 12,31033625 35,00887461 43,44853447 67,80122357 19,64507557 15,865 85,98393272 45,90967463 33,48142645 25,11523737 48,42526115 Cluster Table 6.
Closest Distance of Iteration 5
Region
West Sumba
East Sumba
Kupang
South Central Timor
North Central Timor
Belu
Alor
Lembata
East Flores Sikka
End
Ngada
Manggarai Rote Ndao
West Manggarai Central Sumba
Southwest Sumba
Nagekeo
East Manggarai Sabu Raijua
Kupang City 125,7396813 28,20539627 185,952243 169,6011332 141,3963542 130,3035146 219,8167203 14,96939679 114,8219615 12,31033625 35,00887461 43,44853447 67,80122357 19,64507557 15,865 85,98393272 45,90967463 33,48142645 25,11523737 48,42526115 507,084508 27,77974596 139,4325234 37,37264087 22,48491558 53,21116961 57,93962375 71,52621897 166,4310239 38,87893297
161,3186731
120,7234147
109,1034895
88,41056821
167,8447412
166,3206885
134,0959145
195,0919197
185,3972569
176,5017199
198,421269
355,5002512
381,5193311
491,3318634
321,7452408
337,8579583
382,3519321
385,8134264
294,4401467
521,4019563
393,9390816
515,4716287
475,9936974
464,1379105
440,9920634
521,6895629
520,6755228
476,8658092
548,6647428
539,9833331
531,602295
551,8532414
After the calculation process is completed, it can be grouped into areas with low crime (C.
, areas with medium crime (C.
and areas with high crime (C.
based on the application of the K-Means algorithm.
The following can be seen in table 7, table 8, and table 9 for the grouping of types of crimes against human physicality.
Rizkah Fadillah.
JurIKTI | Page 7 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti Table 7.
Areas with Low Crime (C.
Region East Sumba Lembata Sikka Ende Ngada Manggarai Rote Ndao West Manggarai Central Sumba Southwest Sumba Nagekeo East Manggarai Sabu Raijua Table 8.
Regions with Moderate Crimes (C.
Region West Sumba Kupang South Central Timor North Central Timor Belu Alor East Flores Table 9.
Areas with High Crime (C.
Region Kupang City Based on the data in table 7, it can be seen that the group of areas with low crime (C.
totals 13 regions, the group of areas with medium crime (C.
amounts to 7 regions, and the group of areas with high crime (C.
amounts to 1 So that from this grouping, the government or the authorities must pay more attention to areas based on their crime groups to reduce cases of crimes against human beings.
3 Visualization In this study, we used rapidminer to visualize the K-Means clustering method.
RapidMiner is a data science platform that allows users to perform data mining, text mining, and predictive analysis with more than 500 data mining operators.
RapidMiner allows users to design and run analytical pipelines automatically, has powerful data processing capabilities, is available in a free version for students and early users, is compatible with various operating systems, and can be used for prototyping, application development, and research.
The following can be seen in figure 3 of the operator input process used in determining the number of clusters to be formed.
Figure 3.
RapidMiner Operator Input Proser Based on figure 3, it can be seen in the process using the Read Excel and K-Means operators.
Read Excel aims to read the data to be processed and K-Means aims to classify the data, in the K-Means operator determines the K parameters, in this study the K parameters are determined as much as K=3.
Once all the operators are well connected, then run the process and the results can be seen in table 10.
Rizkah Fadillah.
JurIKTI | Page 8 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti Table 10.
K-Means Classification Results with RapidMiner Region
West Sumba
East Sumba
Kupang
South Central Timor
North Central Timor
Belu
Alor
Lembata
East Flores Sikka
End
Ngada
Manggarai Rote Ndao
West Manggarai Central Sumba
Southwest Sumba
Nagekeo
East Manggarai Sabu Raijua
Kupang City Number of Classifications of Types of Crimes Against Human Physique Murder
Rape
Normal/Mild Persecution Severe Persecution Abduction
2019 2020 2019 2020
Cluster Figure 4.
Results of Bar Chart Visualization Figure 5.
Results of Visualization of K-Means with RapidMiner Figure 6.
Results Cluster_0 Rizkah Fadillah.
JurIKTI | Page 9 Jurnal Ilmu Komputer.
Teknologi Dan Informasi Vol.
No.
Januari 2025.
Hal 1 - 11 ISSN: 2963-0169 (Online - Elektroni.
https://journal.
id/index.
php/jurikti Gambar 7.
Hasil Cluster_1 Gambar 8.
Hasil Cluster_2 4 Result Analysis The last stage is the analysis of the results, where in this study we use the K-Means method for the grouping of types of crimes against the human body in the East Nusa Tenggara region by performing manual calculations using the Euclidean equation formula (Euclidean Distanc.
to determine the closest distance from each data to the centroid.
This study also carried out visualization using rapidminer applications, from manual calculations and rapidminer the results were different cluster arrangements but the areas grouped were the same, namely, the results of manual calculations C1 there were 13 regions.
C2 there were 7 regions and C3 there was 1 region.
The results of the rapidminer Cluster_0 there are 7 regions.
Cluster_1 there are 1 regions and Cluster_2 there are 13 regions.
The regions that are divided into groupings using rapidminer or manual calculation have the same results and there is no difference in region.
CONCLUSION
The conclusion that can be reached from this study after applying the K-Means algorithm in classifying the types of crimes against human bodies in 2019 to 2020 in the East Nusa Tenggara region is that there are 3 centroids.
C1 for areas with low crime.
C2 for areas with moderate crime and C3 for areas with high crime.
The initial centroid value is determined randomly and then for the next centroid is adjusted to the result of the calculation of the closest distance .
The final results obtained are areas with low crime totaling 13 regions, namely East Sumba.
Lembata.
Sikka.
Ende.
Ngada.
Manggarai.
Rote Ndao.
West Manggarai.
Central Sumba.
Southwest Sumba.
Nagekeo.
East Manggarai, and Sabu Raijua.
There are 7 areas with moderate crimes, namely West Sumba.
Kupang.
South Central Timor.
North Central Timor.
Belu.
Alor, and East Flores.
As for the area with high crime, there is 1 area, namely Kupang City.
REFERENCES