International Journal Of Artificial Intelegence Research ISSN: 2579-7298

Vol 6, No 1, June 2022

K-Means and K-NN Methods For Determining Student Interest

Guslendraa1*, Sarjon Defit a,2
aPutra Indonesia University "YPTK" Padang, Indonesia
1
[email protected]
* corresponding author

ARTICLE INFO ABSTRACT

Article history: Putra Indonesia University 'YPTK' Padang's Department of Information
Received 22 Ags 2021 Systems, Faculty of Computing Science has three specializations, namely
Revised 25 Feb 2022 Information Technology Management, Business Information Systems and
Accepted 28 May 2022 Industrial Information Systems. In the fifth semester, the acquisition of
specializations takes place. In the next semester the selection of specialist
programs will be determined. The option of the degree is adapted to students'
needs and capacities. The acquisition of results generated in the previous
semester can be seen. The objective of this survey is to provide students with
suggestions for collection of degrees. The study was performed using K-
Keywords:
Means and K-Nearest Neighbor methods in order to obtain classification of
K-Mean,
students and the correlation between recent cases and past cases. This analysis
K-NN,
uses 13 characteristics, of which 12 are predictors and 1 is the option. The test
Specialization,
results can be used as a way to suggest the student preferences based on preset
Student
attributes through the K-Means and K-NN methods
Copyright © 2017 International Journal of Artificial Intelegence Research.
All rights reserved.

I. Introduction
Each university would like all graduates to know their fields. Graduates' knowledge depends
heavily on the suitability of students' preferences and talents. In three fields of specialization,
namely Information Technology Management, Business Information Systems and Industrial
Information System, the YPTK University Informatics Study Program produces graduates. Starting
from the 5th semester, the selection of specializations may be made. In the next semester, the
selected specialty will decide the classes that will determine the student's area of expertise. It is
dependent on many subjects in order to assess the best option of specialisation. For three specialized
fields, 12 topics form the foundation of expertise, where the four subjects are each focused. This
does not, of course, entail reducing the expense of other courses to be offered to pupils.
Research in the selection of concentration has also been carried out by [1] on the Yogyakarta
Amikom International Class Students using the K-Means method which produces suitable
concentrations based on the scores of several subjects. In contrast to the current research,
determining the suitable concentration is determined from the results of the cluster produced by the
K-Means method.
In another study [2] conducted a study to group the Final Semester Student Concentration Course
Classes at Ichsan Gorontalo University using the K-Means and KMeans KNN methods. This study
produces clusters to classify Concentration Course Classes for final semester students and each of
these clusters has a predictive value for the two clusters.
Grouping and prediction data with the K-Means method and K-NN has also been done by [3] to
produce a decision-making tool giving rewards to the employees of the UPI Convention Group, [4]
in the Determination Regional Priority Services Birth Certificate Bogor, [5] in his research
Implementation method K-Nn To Predict Results Agriculture in Malang, [6] on image Segmentation
in wheezing the identification of diseases and diseases, [7] on the Acceptance of New Technology-
minded Teacher Candidates and Administrative Employees, and [8] on Sentiment Analysis for
Travel Agent Reviews.

DOI: 10.29099/ijair.v6i1.222 W : http://ijair.id | E : [email protected]

 International Journal Of Artificial Intelegence Research ISSN: 2579-7298
Vol 6, No 1, June 2022

In making this specialization selection, students often hesitate in making decisions because of
their ignorance and the influence of other friends so that it is often not following their abilities. This
will also have an impact on the quality of graduates from the study program.
This study aims to provide recommendations to students in determining their specialization based
on the course scores they have obtained in the previous semester. With this research, it is expected
that the chosen students will choose according to their abilities so that it will improve the quality of
graduates.

II. Methods
Several phases of activity are carried out in this research. The operation starts from the collection
of data. The data collected would be pre-processed in order to acquire training data and test data for
analysis. Via the grouping process, training data can be performed. The classification is performed
using the system K-Means [9][10]. The results of the classification yield clusters of student data that
are recommended to identify and not recommended specializations. This is achieved on the basis of
the previous values which later form the basis of the course. In order to generate a recommended
speciality for students, the recommended data are analyzed using the K-Nearest neighbor system
[11][12][13][14][15] and test data. Figure 1 shows the mechanism for this processing of data.

Figure 1 Data Processing Framework

The framework for the above data processing can be explained as follows:
A. Preprocessing
This is the first step taken before processing the data mining process. It is done to eliminate
distractions that exist in the data [16][17]. Poor data can be caused by several things, namely:
1. Students have not taken all the subjects that are the basis for determining specialization.
2. Some scores fail either one subject or more than the basic subject of specialization (letter
E-grade )
3. Subject data taken are basic courses of specialization in semesters one to four.
B. Training Data
Training data is pre-existing data based on facts that have already occurred. In this study, the
training data were taken from the student data of each specialization with the value of the basic
subject being the maximum value.
C. Test Data
Test data is data that is already valuable which is used to calculate the accuracy of the formed
classification model. In this study, the test data were taken from the results of the student data
grouping process that had been carried out using the K-Means method.
D. K-Means
The K-Means algorithm is an iterative grouping of data sets into some predefined
clusters [10]. The K-Means method is data grouping by maximizing the similarity of data in
one cluster and minimizing the similarity of data between clusters [18]. The measure of similarity
used in clusters is a function of distance. So that maximization of data similarity is obtained based
on the shortest distance between the data and the centroid point [19]. The K-Means method is
a method used to classify student data to obtain the recommended data for selecting specializations
and those that have not. The use of the K-Means method is to group data into some clusters by
comparing the distance between the data and the centroid [20][21]. Algoritma K-Means is not
influenced by the order of objects. An important step in using the K-means method is to determine

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
ISSN: 2579-7298 International Journal Of Artificial Intelegence Research
Vol 6, No 1, June 2022

the centroid, the number of clusters, and the distance of the centroid. By forming
several clusters using the K-Means algorithm, you can also find out the distance
between the central cluster ( centroid ) on the data to be analyzedIn this study, the determination of
the initial value of the centroid was carried out randomly [1][22]. To determine the distance between
the data and the centroid, the following formula is used:

The process will be carried out until there is no change in the distance between the data and the
cluster center. The results of this data grouping process will be used in the next process.
E. Recommended Data
Recommendation data is data generated from the data grouping process using the K-Means
method. The recommended data is data that is complete and has taken all the basic courses of the
specialization.
F. Data Not Recommended
Data not recommended is data resulting from the grouping process that has not met the
criteria. This could be since there are still some scores that are obtained below the passing standards,
especially those for the basic subjects of specialization.
G. K-Nearest Neighbor (K-NN)
K-NN is an algorithm that classifies new objects based on attributes and training
samples [23][24]. KNN is a non-parametric lazy learning algorithm because it does not make any
assumptions on the distribution of the main data [25][26]. This algorithm works based on the
shortest distance from the query instance to the training data [27][28][29]. This algorithm uses a
supervised algorithm where the results of the new test samples are grouped based on the majority of
the categories on the K-NN [30]. The results of the grouping process carried out in the previous
process will be processed with the K Nearest Neighboard method to predict what suitable
specializations are taken by a student by looking at the similarity value generated using the
following formula [3] :

Where T is the new case, S is the storage case, n is the number of attributes in each case, i is the
individual attribute from 1 to n and wi is the attribute feature weight. In this process, the attributes to
be used will be determined. In this case, there are 12 attributes used which are taken from the
courses which are the basis for each specialization.
H. Specialization Recommendations
The specialization recommendation is the result of a process that has been carried out using the
K-NN method. The recommended specialization is that which has the highest similarity value
obtained from the closeness of the value between the learning data and the test data

III. Result and Discussion

This study uses data derived from student scores for the Information Systems Study Program at
Putra Indonesia University 'YPTK' Padang. The Information Systems Study Program has three
specializations that will determine the area of expertise of the graduate. In determining this
specialization, students often experience doubts, resulting in students determining their
specialization that is not according to their abilities, this can be seen from the scores they obtained in
the previous semester. Errors in the selection of specialization will greatly affect the quality of
graduates. For this reason, it is necessary to group students who will determine their
specialization. The K-Means method will be used to group students into two groups. The first group
are students who are recommended to determine specialization and the second group are students
who have not been recommended to determine specialization. The process of grouping is carried out
by considering several subjects which are the basis for each specialization. There are 12 subjects and

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
 International Journal Of Artificial Intelegence Research ISSN: 2579-7298
Vol 6, No 1, June 2022

3 specializations that have been assigned to be used in this classification process as shown in Table 1
and Table 2.
Table 1 List Of Subjection Basic Subjection

No. Course Name Code

1. Introduction to Management pmnj
2. Introduction to Information Technology pti
3. General Organization Theory tou
4. Management Information System sim
5. Information Systems Concept ksi
6. Business Knowledge pb
7. Accounting information system drain
8. Information System Design Analysis 1 apsi1
9. Algorithms and Data Structures 1 alg1
10. Algorithms and Data Structures 2 alg2
11. Database System sbd
12. Database Design pbd

Table 2 Specialization

No. Name of Specialization

1. Management Information System
2. Business Information Systems
3. Industrial Information Systems
The sample data of the Information System students at the Faculty of Computer Science, Putra
Indonesia University YPTK Padang used can be seen in table 3.
Table 3 One, Two, Three, And Four Semester Student Scores
Information Technology Industrial Information
Business Information Systems
Student name Management Systems
pmnj pti tou sim ksi pb drain apsi1 alg1 alg2 sbd pbd
Abidzar Ghifari Zandra A A B A A B A A A A A A
Afdila Zartika A B A B A A A A B A A A
Aisyah Nurminas B B B B A A A A B B B A
Amhar Al unawar B B C C B A C D C D C C
Angga Agustiadi B B B B B B B C A C B C
Ari Suhanda D B C B C C C C A A A A
Ayu Winanda B B B B B B B B B C B B
Deviani A B B B A A B B A B B A
Dino Febrian Doni B B B C A B B D B B B D
Edi Susilo B A B B B A B B B C B A
Elri Suhendra B B B B B B C D B D C B
Fadila Erina B B A A A A B B B C B B
Heru Pramana Firnu C C B B C B C B A C B C
Inspiration of the Son
D C C B C B C C B D D D
of Gifts
Indri Yani Putri B A A A A A A A A B B A
Kelvin Frendinata C C B B C A C C C D B A
Leonida Cipta
A B B B A A B B A B B A
Meidayanti
Malfo Dewo A A A A A A A A A A A A
Mohd. Martha. M B D B B B B C D B B B D
Muhammad Fadel B B B B B B C C B C B B
Muhammad Reki
C B B B B C B D C D B C
Andika
Nada Permata Sari B B B B B A B B B D B B
Nadya Dwi Yasra B B B B A A A B B A B B
Nur Farahana B B B A B A A B B B B B
Nurul Indah Azizah A A A A A A A B A B A A
Panji Patikawa B B B B C B A C A B B B

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
ISSN: 2579-7298 International Journal Of Artificial Intelegence Research
Vol 6, No 1, June 2022

Qaidah Pratari
A A A A B A B B B B B A
Rahmatullah
Randi Sulaeman B B B B A A B A B A C A
Renza Nazirah Faiqah B B A B A A B A B B B B
Rika Ridla Juita B B B B B B B B B B B B
Ririn Uswatun H B A B B A A B B B B B A
Rizki Permana Putra D C D B C A D D D C B C
Rudi Greetings B C C B B B D D B C C D
Shinta Amelia Ananda B B B A A A B D C B B A
Sukri Azhari C B B B C B B D C C C C
Syamsuri Sani B B C B B A D D B C C C
Teguh Arisandi C B B B C B B B B D C C
Thia Pratama Tanjung B B B D B A D D D C C D
Tri Wanda Yasman B B B B B A C D B D B C
Viki Pratama C B B B C C C D B C B B
Wahyu Ferdiant B B B B B C B B A C D B
Windika Pilyadi B A B B A A C D A A C D
Yogi Fajr Bahari B A B A A A A A B B A B

The process of grouping values with K-Means is a process of numerical values. So that the data
above needs a value transformation process so that all natural values are in the form of numbers. The
transformation is made with the values A = 4, B = 3, C = 2 and D = 1. The results of the value
transformation to be processed can be seen in table 4
Table 4 .Transformation Of Value For One, Two, Three, And Four Semester Students
Information Technology Business Information Industrial Information
Student name Management Systems Systems
pmnj pti tou sim ksi pb drain apsi1 alg1 alg2 sbd Pbd

Abidzar Ghifari Zandra 4 4 3 4 4 3 4 4 4 4 4 4

Afdila Zartika 4 3 4 3 4 4 4 4 3 4 4 4

Aisyah Nurminas 3 3 3 3 4 4 4 4 3 3 3 4

Amhar Al Munawar 3 3 2 2 3 4 2 1 2 1 2 2

Angga Agustiadi 3 3 3 3 3 3 3 2 4 2 3 2

Ari Suhanda 1 3 2 3 2 2 2 2 4 4 4 4

Ayu Winanda 3 3 3 3 3 3 3 3 3 2 3 3

Deviani 4 3 3 3 4 4 3 3 4 3 3 4

Dino Febrian Doni 3 3 3 2 4 3 3 1 3 3 3 1

Edi Susilo 3 4 3 3 3 4 3 3 3 2 3 4

Elri Suhendra 3 3 3 3 3 3 2 1 3 1 2 3

Fadila Erina 3 3 4 4 4 4 3 3 3 2 3 3

Heru Pramana Firnu 2 2 3 3 2 3 2 3 4 2 3 2

Inspiration of the Son of Gifts 1 2 2 3 2 3 2 2 3 1 1 1

Indri Yani Putri 3 4 4 4 4 4 4 4 4 3 3 4

Kelvin Frendinata 2 2 3 3 2 4 2 2 2 1 3 4

Leonida Cipta Meidayanti 4 3 3 3 4 4 3 3 4 3 3 4

Malfo Dewo 4 4 4 4 4 4 4 4 4 4 4 4

Mohd. Martha. M 3 1 3 3 3 3 2 1 3 3 3 1

Muhammad Fadel 3 3 3 3 3 3 2 2 3 2 3 3

Muhammad Reki Andika 2 3 3 3 3 2 3 1 2 1 3 2

Nada Permata Sari 3 3 3 3 3 4 3 3 3 1 3 3

Nadya Dwi Yasra 3 3 3 3 4 4 4 3 3 4 3 3

Nur Farahana 3 3 3 4 3 4 4 3 3 3 3 3

Nurul Indah Azizah 4 4 4 4 4 4 4 3 4 3 4 4

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
 International Journal Of Artificial Intelegence Research ISSN: 2579-7298
Vol 6, No 1, June 2022

Panji Patikawa 3 3 3 3 2 3 4 2 4 3 3 3

Qaidah Pratari Rahmatullah 4 4 4 4 3 4 3 3 3 3 3 4

Randi Sulaeman 3 3 3 3 4 4 3 4 3 4 2 4

Renza Nazirah Faiqah 3 3 4 3 4 4 3 4 3 3 3 3

Rika Ridla Juita 3 3 3 3 3 3 3 3 3 3 3 3

Ririn Uswatun Hasanah 3 4 3 3 4 4 3 3 3 3 3 4

Rizki Permana Putra 1 2 1 3 2 4 1 1 1 2 3 2

Rudi Greetings 3 2 2 3 3 3 1 1 3 2 2 1

Shinta Amelia Ananda 3 3 3 4 4 4 3 1 2 3 3 4

Sukri Azhari 2 3 3 3 2 3 3 1 2 2 2 2

Syamsuri Sani 3 3 2 3 3 4 1 1 3 2 2 2

Teguh Arisandi 2 3 3 3 2 3 3 3 3 1 2 2

Thia Pratama Tanjung 3 3 3 1 3 4 1 1 1 2 2 1

Tri Wanda Yasman 3 3 3 3 3 4 2 1 3 1 3 2

Viki Pratama 2 3 3 3 2 2 2 1 3 2 3 3

Wahyu Ferdiant 3 3 3 3 3 2 3 3 4 2 1 3

Windika Pilyadi 3 4 3 3 4 4 2 1 4 4 2 1

Yogi Fajr Bahari 3 4 3 4 4 4 4 4 3 3 4 3

The first step is to determine the data centroid. Centroid is determined randomly. In this process,
2 centroids are selected as shown in Table 5 below:
Table 5 Centroid Table

Centroid Student name pmnj pti tou sim ksi Pb drain apsi1 alg1 alg2 sbd pbd

C1 NADYA DWI YASRA B B B B A A A B B A B B

C2 RIZKI PERMANA PUTRA D C D B C A D D D C B C

The value of the centroid is transformed according to the predefined conditions. The results of
the transformation can be seen in table 6. below.
Table 6 Centroid Transformation

Centroid Student name pmnj pti tou sim ksi Pb drain apsi1 alg1 alg2 sbd pbd
C1 Nadya Dwi Yasra 3 3 3 3 4 4 4 3 3 4 3 3

C2 Rizki Permana Putra 1 2 1 3 2 4 1 1 1 2 3 2

The second step is to determine the distance between the data and its centroid values. Table 7.
The following is the result of the process according to the formula mentioned in the previous
chapter.
Table 7 Data Distance With Centroid Value
Mnj Information Business Information Industrial Information Distance
Score
Student name Technology Systems Systems To
pmnj pti tou sim ksi pb drain apsi1 alg1 alg2 sbd pbd C1 C2 Cluster
Abidzar Ghifari Zandra 4 4 3 4 4 3 4 4 4 4 4 4 2.83 7.68 C1
Afdila Zartika 4 3 4 3 4 4 4 4 3 4 4 4 2.24 7.35 C1
Aisyah Nurminas 3 3 3 3 4 4 4 4 3 3 3 4 1.73 6.32 C1
Amhar Al Munawar 3 3 2 2 3 4 2 1 2 1 2 2 4.80 3.46 C2
Angga Agustiadi 3 3 3 3 3 3 3 2 4 2 3 2 3.16 5.00 C1
Ari Suhanda 1 3 2 3 2 2 2 2 4 4 4 4 4.58 5.10 C1
Ayu Winanda 3 3 3 3 3 3 3 3 3 2 3 3 2.65 4.90 C1
Deviani 4 3 3 3 4 4 3 3 4 3 3 4 2.24 6.32 C1

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
ISSN: 2579-7298 International Journal Of Artificial Intelegence Research
Vol 6, No 1, June 2022

Dino Febrian Doni 3 3 3 2 4 3 3 1 3 3 3 1 3.46 5.00 C1

Edi Susilo 3 4 3 3 3 4 3 3 3 2 3 4 2.83 5.39 C1
Elri Suhendra 3 3 3 3 3 3 2 1 3 1 2 3 4.47 4.36 C2
Fadila Erina 3 3 4 4 4 4 3 3 3 2 3 3 2.65 5.66 C1
Heru Pramana Firnu 2 2 3 3 2 3 2 3 4 2 3 2 4.12 4.47 C1
Inspiration of the Son of
1 2 2 3 2 3 2 2 3 1 1 1 5.74 3.74 C2
Gifts
Indri Yani Putri 3 4 4 4 4 4 4 4 4 3 3 4 2.65 7.35 C1
Kelvin Frendinata 2 2 3 3 2 4 2 2 2 1 3 4 4.69 3.61 C2
Leonida Cipta Meidayanti 4 3 3 3 4 4 3 3 4 3 3 4 2.24 6.32 C1
Malfo Dewo 4 4 4 4 4 4 4 4 4 4 4 4 2.83 7.94 C1
Mohd. Martha. M 3 1 3 3 3 3 2 1 3 3 3 1 4.36 4.24 C2
Muhammad Fadel 3 3 3 3 3 3 2 2 3 2 3 3 3.32 4.24 C1
Muhammad Reki Andika 2 3 3 3 3 2 3 1 2 1 3 2 4.69 4.12 C2
Nada Permata Sari 3 3 3 3 3 4 3 3 3 1 3 3 3.32 4.90 C1
Nadya Dwi Yasra 3 3 3 3 4 4 4 3 3 4 3 3 0.00 5.92 C1
Nur Farahana 3 3 3 4 3 4 4 3 3 3 3 3 1.73 5.48 C1
Nurul Indah Azizah 4 4 4 4 4 4 4 3 4 3 4 4 2.83 7.42 C1
Panji Patikawa 3 3 3 3 2 3 4 2 4 3 3 3 2.83 5.57 C1
Qaidah Pratari Rahmatullah 4 4 4 4 3 4 3 3 3 3 3 4 2.83 6.40 C1
Randi Sulaeman 3 3 3 3 4 4 3 4 3 4 2 4 2.00 6.24 C1
Renza Nazirah Faiqah 3 3 4 3 4 4 3 4 3 3 3 3 2.00 6.08 C1
Rika Ridla Juita 3 3 3 3 3 3 3 3 3 3 3 3 2.00 5.00 C1
Ririn Uswatun Hasanah 3 4 3 3 4 4 3 3 3 3 3 4 2.00 5.74 C1
Rizki Permana Putra 1 2 1 3 2 4 1 1 1 2 3 2 5.92 0.00 C2
Rudi Greetings 3 2 2 3 3 3 1 1 3 2 2 1 5.10 3.61 C2
Shinta Amelia Ananda 3 3 3 4 4 4 3 1 2 3 3 4 3.00 4.90 C1
Sukri Azhari 2 3 3 3 2 3 3 1 2 2 2 2 4.24 3.61 C2
Syamsuri Sani 3 3 2 3 3 4 1 1 3 2 2 2 4.58 3.46 C2
Teguh Arisandi 2 3 3 3 2 3 3 3 3 1 2 2 4.24 4.58 C1
Thia Pratama Tanjung 3 3 3 1 3 4 1 1 1 2 2 1 5.57 4.00 C2
Tri Wanda Yasman 3 3 3 3 3 4 2 1 3 1 3 2 4.36 4.00 C2
Viki Pratama 2 3 3 3 2 2 2 1 3 2 3 3 4.58 4.00 C2
Wahyu Ferdiant 3 3 3 3 3 2 3 3 4 2 1 3 3.87 6.00 C1
Windika Pilyadi 3 4 3 3 4 4 2 1 4 4 2 1 3.87 5.66 C1
Yogi Fajr Bahari 3 4 3 4 4 4 4 4 3 3 4 3 2.24 6.48 C1

The third step repeats the process by determining the new centroid value by calculating the
centroid of each cluster for each attribute until the position of the cluster does not change.
The new centroid values can be seen in table 8 below:
Table 8 Iteration Centroid Value 2.

Cluster 1 3.10 3.27 3.20 3.27 3.43 3.57 3.20 2.93 3.37 2.87 3.03 3.23
Cluster 2 2.38 2.54 2.54 2.77 2.62 3.31 1.85 1.15 2.38 1.62 2.38 2.00

Repetition will be carried out until there is no more change in the Cluster value. In this study,
there were 4 repetitions of the process in determining the classification of student data.From the
results of the process, there were two groups of students, Group C1 was advised to take
specialization while group C2 was not recommended. The results of this grouping can be seen in the
following tables 9 and 10.
Table 9 Group C1
Information
Business Information Industrial Distance
Technology Cluster
No. Student name Systems Information Systems To
Management
pmnj pti tou sim ksi pb drain apsi1 alg1 alg2 sbd pbd C1 C2
1 Abidzar Ghifari Zandra 4 4 3 4 4 3 4 4 4 4 4 4 2.38 5.53 C1
2 Afdila Zartika 4 3 4 3 4 4 4 4 3 4 4 4 2.21 5.30 C1
3 Aisyah Nurminas 3 3 3 3 4 4 4 4 3 3 3 4 1.44 4.36 C1
4 Ari Suhanda 1 3 2 3 2 2 2 2 4 4 4 4 4.18 4.25 C1
5 Ayu Winanda 3 3 3 3 3 3 3 3 3 2 3 3 1.65 2.35 C1

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
 International Journal Of Artificial Intelegence Research ISSN: 2579-7298
Vol 6, No 1, June 2022

6 Deviani 4 3 3 3 4 4 3 3 4 3 3 4 1.44 3.94 C1

7 Edi Susilo 3 4 3 3 3 4 3 3 3 2 3 4 1.60 3.28 C1
8 Fadila Erina 3 3 4 4 4 4 3 3 3 2 3 3 1.75 3.25 C1
9 Indri Yani Putri 3 4 4 4 4 4 4 4 4 3 3 4 1.87 5.02 C1
10 Leonida Cipta Meidayanti 4 3 3 3 4 4 3 3 4 3 3 4 1.44 3.94 C1
11 Malfo Dewo 4 4 4 4 4 4 4 4 4 4 4 4 2.41 5.71 C1
12 Nada Permata Sari 3 3 3 3 3 4 3 3 3 1 3 3 2.32 2.59 C1
13 Nadya Dwi Yasra 3 3 3 3 4 4 4 3 3 4 3 3 1.60 3.88 C1
14 Nur Farahana 3 3 3 4 3 4 4 3 3 3 3 3 1.38 3.41 C1
15 Nurul Indah Azizah 4 4 4 4 4 4 4 3 4 3 4 4 2.04 5.02 C1
16 Panji Patikawa 3 3 3 3 2 3 4 2 4 3 3 3 2.39 3.10 C1
17 Qaidah Pratari Rahmatullah 4 4 4 4 3 4 3 3 3 3 3 4 1.70 4.14 C1
18 Randi Sulaeman 3 3 3 3 4 4 3 4 3 4 2 4 2.02 4.40 C1
19 Renza Nazirah Faiqah 3 3 4 3 4 4 3 4 3 3 3 3 1.58 3.83 C1
20 Rika Ridla Juita 3 3 3 3 3 3 3 3 3 3 3 3 1.32 2.62 C1
21 Ririn Uswatun Hasanah 3 4 3 3 4 4 3 3 3 3 3 4 1.22 3.69 C1
22 Shinta Amelia Ananda 3 3 3 4 4 4 3 1 2 3 3 4 2.78 3.40 C1
23 Wahyu Ferdiant 3 3 3 3 3 2 3 3 4 2 1 3 3.15 3.24 C1
24 Yogi Fajr Bahari 3 4 3 4 4 4 4 4 3 3 4 3 1.89 4.58 C1

Table 10 Group C2
Information Technology Business Information Industrial Distance
Cluster
No. Student name Management Systems Information Systems To
pmnj pti tou sim ksi pb Drain apsi1 alg1 alg2 sbd pbd C1 C2

1 Amhar Al Munawar 3 3 2 2 3 4 2 1 2 1 2 2 4.49 1.93 C2

2 Angga Agustiadi 3 3 3 3 3 3 3 2 4 2 3 2 2.56 1.95 C2

3 Dino Febrian Doni 3 3 3 2 4 3 3 1 3 3 3 1 3.79 2.53 C2

4 Elri Suhendra 3 3 3 3 3 3 2 1 3 1 2 3 3.66 1.72 C2

5 Heru Pramana Firnu 2 2 3 3 2 3 2 3 4 2 3 2 3.49 2.42 C2

6 Inspiration of the Son of Gifts 1 2 2 3 2 3 2 2 3 1 1 1 5.45 2.82 C2

7 Kelvin Frendinata 2 2 3 3 2 4 2 2 2 1 3 4 3.89 2.83 C2

8 Mohd. Martha. M 3 1 3 3 3 3 2 1 3 3 3 1 4.45 2.48 C2

9 Muhammad Fadel 3 3 3 3 3 3 2 2 3 2 3 3 2.41 1.57 C2

10 Muhammad Reki Andika 2 3 3 3 3 2 3 1 2 1 3 2 4.24 2.18 C2

11 Rizki Permana Putra 1 2 1 3 2 4 1 1 1 2 3 2 5.80 3.35 C2

12 Rudi Greetings 3 2 2 3 3 3 1 1 3 2 2 1 4.86 1.97 C2

13 Sukri Azhari 2 3 3 3 2 3 3 1 2 2 2 2 4.00 1.72 C2

14 Syamsuri Sani 3 3 2 3 3 4 1 1 3 2 2 2 4.18 1.76 C2

15 Teguh Arisandi 2 3 3 3 2 3 3 3 3 1 2 2 3.56 2.34 C2

16 Thia Pratama Tanjung 3 3 3 1 3 4 1 1 1 2 2 1 5.57 3.12 C2

17 Tri Wanda Yasman 3 3 3 3 3 4 2 1 3 1 3 2 3.74 1.54 C2

18 Viki Pratama 2 3 3 3 2 2 2 1 3 2 3 3 3.87 2.07 C2

19 Windika Pilyadi 3 4 3 3 4 4 2 1 4 4 2 1 4.12 3.44 C2

The data in the C1 group will be further processed to determine the appropriate specialization for
these students. The process will be carried out using the K Nearest Neighbor method to produce a
similarity value with the test data used.

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
ISSN: 2579-7298 International Journal Of Artificial Intelegence Research
Vol 6, No 1, June 2022

In this process, the attributes that will be used will be determined where there are 12 attributes
taken from the course which are the basis for each specialization. Each attribute is given a symbol
and weighted according to its level of need. The attribute names, symbols, and weighting can be
seen in table 11. below:
Table 11 Name Attributes And Symbols

No. (Symbol) Attribute Name Weight

1 (A) Introduction to Management 0.9
2 (B) Introduction to Information Technology 0.9
3 (C) General Organization Theory 0.9
4 (D) Management Information Systems 0.9
5 (E) Information Systems Concepts 0.9
6 (F) Business Knowledge 0.9
7 (G) Accounting Information Systems 0.9
8 (H) Analysis of Information System Design 0.9
9 (I) Algorithms and Data Structures 1 0.9
10 (J) Algorithms and Data Structures 2 0.9
11 (K) Database Systems 0.9
12 (L) Database Design 0.9

Each attribute has an attribute value which can be seen in table 12. below:
Table 12 Attributes Value
No. Attribute Attribute Value No. Attribute Attribute Value
(G) Accounting Information
1
(A) Introduction to Management (A1) Grade A. 7 Systems (G1) Value A.
(A2) Value of B. (G2) Value of B
(A3) Value of C (G3) Value of C
(A4) Value of D (G4) Value of D
(B) Introduction to Information (H) Analysis of Information
2
Technology (B1) Grade A. 8 System Design (H1) Value A
(B2) Value of B (H2) Value B
(B3) Value of C (H3) Value of C
(B4) Value of D (H4) Value of D
(I) Algorithms and Data
3
(C) General Organization Theory (C1) Grade A. 9 Structures 1 (I1) Grade A.
(C2) Value of B (I2) Value of B
(C3) Value of C (I3) Value of C
(C4) Value of D (I4) Value of D
(J) Algorithms and Data
4
(D) Management Information Systems (D1) Grade A. 10 Structures 2 (J1) Grade A.
(D2) Value of B (J2) Value of B
(D3) Value of C (J3) Value of C
(D4) Value of D (J4) Value of D
5 (E) Information Systems Concepts (E1) Grade A. 11 (K) Database Systems (K1) Grade A.
(E2) Value of B (K2) Value of B
(E3) Value of C (K3) Value of C
(E4) Value of D (K4) Value of D
(F) Business Knowledge (F1) Value A 12 (L) Database Design (L1) Grade A.
(F2) Value of B (L2) Value of B
(F3) Value of C (L3) Value of C
(F4) Value of D (L4) Value of D

The next process is to calculate the proximity of the attribute values for all the attributes
used. Tables 13. and 14. The following are examples of the calculation process in finding the value
of the proximity of each Attribute Value.

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
 International Journal Of Artificial Intelegence Research ISSN: 2579-7298
Vol 6, No 1, June 2022

Table 13 Classification (A) Introduction To Management

No. Attribute Value Score Weight

1 (A1) Grade A. 4
2 (A2) Value of B. 3
0.9
3 (A3) Value of C 2
4 (A4) Value of D 1

Table 14 Approach To Attributes Value (A) Introduction To Management

A1 A2 A3 A4
A1 1 0.75 0.5 0.25
A2 0.75 1 0.67 0.33
A3 0.5 0.67 1 0.5
A4 0.25 0.33 0.5 1

The next step is to select sample data for training which can be seen in table 15. below:
Table 15 Training Data Sample
No. Name A B C D E F G. H I J K L Decision
1 Afdila Zartika A1 B2 C1 D2 E1 F1 G1 H1 I2 J1 K1 L1 Business Information Systems
2 Nurul Indah Azizah A1 B1 C1 D1 E1 F1 G1 H2 I1 J2 K1 L1 Information Technology Management
3 Abidzar Ghifari Zandra A1 B1 C2 D1 E1 F2 G1 H1 I1 J1 K1 L1 Industrial Information Systems
4 Aisyah Nurminas A2 B2 C2 D2 E1 F1 G1 H1 I2 J2 K2 L1 Business Information Systems
5 Qaidah Pratari Rahmatullah A1 B1 C1 D1 E2 F1 G2 H2 I2 J2 K2 L1 Information Technology Management
6 Ari Suhanda A4 B2 C3 D2 E3 F3 G3 H3 I1 J1 K1 L1 Industrial Information Systems

The testing data used can be seen in Table 16 below:

Table 16 Testing Data

No. Name A B C D E F G. H I J K L Decision.

7 Leonida Cipta Meidayanti A1 B2 C2 D2 E1 F1 G2 H2 I1 J2 K2 L1

The next step is to calculate the closeness of the new case in Table 16. and the old case Table 15.
The closeness of the new case to case number 1 in the old case is shown in Table 17 below:
Table 17 New Case Approach With Old Case.

No. Attribute Case 1 New Case Proximity (s) Weight (w) S*W

1 A A1 A1 1 0.9 0.9

2 B B2 B2 1 0.9 0.9

3 C C1 C2 0.75 0.9 0.675

4 D D2 D2 1 0.9 0.9

5 E E1 E1 1 0.9 0.9

6 F F1 F1 1 0.9 0.9

7 G. G1 G2 0.75 0.9 0.675

8 H H1 H2 0.75 0.9 0.675

9 I I2 I1 0.75 0.9 0.675

10 J J1 J2 0.75 0.9 0.675

11 K K1 K2 0.75 0.9 0.675

12 L L1 L1 1 0.9 0.9
10.8 9.45

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
ISSN: 2579-7298 International Journal Of Artificial Intelegence Research
Vol 6, No 1, June 2022

From Table 17, the closeness of the new case to case number 1 can be calculated as follows:
Similarity = 0.9 + 0.9 + 0.675 + 0.9 + 0.9 + 0.9 + 0.675 + 0.675 + 0.675 + 0.675 + 0.675 + 0.9
0.9 + 0.9 + 0.9 + 0.9 + 0.9 + 0.9 + 0.9 + 0.9 + 0.9 + 0.9 + 0.9 + 0.9
= 0.875
The same process is carried out on all training data so that the highest similarity value is obtained
for new cases. The results of the calculation process in this study were tested using data mining
applications. The test results prove that the results of the calculation process have the same value as
the results of the application.
Prediction With K-Nearest Neighbors

IV. Conclusion
This research was conducted by taking training data and test data from the student data of the
Putra Indonesia University Information System 'YPTK' Padang. This study uses 12 attributes taken
from existing courses in semesters up to 4. This research has been able to provide recommendations
for students who will choose specializations under the basic courses of specialization they have
mastered, this can be seen from the results of the course scores obtained. From the results of tests
carried out using Rapid Miner, the calculation results of the methods used have similarities. This
proves that the K-Means and K-NN methods can be used in determining student interest. This
research can also be developed using other methods to get maximum results. Research can also be
carried out to determine the direction of student research based on the grades they got in previous
semesters

References
[1] J. Aranda and W. A. G. Natasya, “Penerapan Metode K-Means Cluster Analysis Pada
Sistem Pendukung Keputusan Pemilihan Konsentrasi Untuk Mahasiswa International Class
Stmik Amikom Yogyakarta,” Semnasteknomedia Online, vol. 4, no. 1, pp. 4-2–1, 2016,
[Online]. Available:
https://ojs.amikom.ac.id/index.php/semnasteknomedia/article/view/1293.

[2] S. Rustam and H. Annur, “Akademik Data Mining (Adm) K-Means Dan K-Means K-Nn
Untuk Mengelompokan Kelas Mata Kuliah Kosentrasi Mahasiswa Semester Akhir,” Ilk. J.
Ilm., vol. 11, no. 3, pp. 260–268, 2019, doi: 10.33096/ilkom.v11i3.487.260-268.

[3] E. Praja, W. Mandala, M. Ridwan, and E. Putri, “DATA MINING PEMBERIAN

REWARD PADA KARYAWAN UPI,” pp. 37–44, 2019.

[4] A. M. M. Anwar, P. Harsani, and A. Maesya, “Penentuan Daerah Prioritas Pelayanan Akta
Kelahiran Dengan Metode K-Nn Dan K-Means,” Komputasi J. Ilm. Ilmu Komput. dan Mat.,
vol. 17, no. 1, pp. 319–328, 2020, doi: 10.33751/komputasi.v17i1.1884.

[5] A. Pungky, “Penerapan metode k-nn untuk memprediksi hasil pertanian di kabupaten
malang,” vol. 3, no. 1, pp. 235–242, 2019.

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
 International Journal Of Artificial Intelegence Research ISSN: 2579-7298
Vol 6, No 1, June 2022

[6] F. G. Febrinanto, C. Dewi, and A. T. Wiratno, “Implementasi Algoritme K-Means Sebagai

Metode Segmentasi Citra Dalam Identifikasi Penyakit Daun Jeruk,” J. Pengemb. Teknol.
Inf. dan Ilmu Komput. Univ. Brawijaya, vol. 2, no. 11, pp. 5375–5383, 2018.

[7] N. N. Dzikrulloh and B. D. Setiawan, “Penerapan Metode K – Nearest Neighbor ( KNN )

dan Metode Weighted Product ( WP ) Dalam Penerimaan Calon Guru Dan Karyawan Tata
Usaha Baru Berwawasan Teknologi ( Studi Kasus : Sekolah Menengah Kejuruan
Muhammadiyah 2 Kediri ),” Pengembangan Teknologi Informasi dan Ilmu Komputer, vol.
1, no. 5. pp. 378–385, 2017.

[8] S. Ernawati and R. Wati, “Penerapan Algoritma K-Nearest Neighbors Pada Analisis
Sentimen Review Agen Travel,” J. Khatulistiwa Inform., vol. VI, no. 1, pp. 64–69, 2018,
[Online]. Available:
https://ejournal.bsi.ac.id/ejurnal/index.php/khatulistiwa/article/view/3802/2626.

[9] I. Kamila, U. Khairunnisa, and M. Mustakim, “Perbandingan Algoritma K-Means dan K-

Medoids untuk Pengelompokan Data Transaksi Bongkar Muat di Provinsi Riau,” J. Ilm.
Rekayasa dan Manaj. Sist. Inf., vol. 5, no. 1, p. 119, 2019, doi: 10.24014/rmsi.v5i1.7381.

[10] F. Nur, M. Zarlis, and B. B. Nasution, “Penerapan Algoritma K-Means Pada Siswa Baru
Sekolahmenengah Kejuruan Untuk Clustering Jurusan,” InfoTekJar (Jurnal Nas. Inform.
dan Teknol. Jaringan), vol. 1, no. 2, pp. 100–105, 2017, doi: 10.30743/infotekjar.v1i2.70.

[11] C. Paramita, E. Hari Rachmawanto, C. Atika Sari, and D. R. Ignatius Moses Setiadi,
“Klasifikasi Jeruk Nipis Terhadap Tingkat Kematangan Buah Berdasarkan Fitur Warna
Menggunakan K-Nearest Neighbor,” J. Inform. J. Pengemb. IT, vol. 4, no. 1, pp. 1–6, 2019,
doi: 10.30591/jpit.v4i1.1267.

[12] R. Enggar Pawening, W. Ja, and far Shudiq, “KLASIFIKASI KUALITAS JERUK LOKAL
BERDASARKAN TEKSTUR DAN BENTUK MENGGUNAKAN METODE k-
NEAREST NEIGHBOR (k-NN),” Ejournal.Unuja.Ac.Id, vol. 1, no. 1, pp. 10–17, 2020,
[Online]. Available: http://ejournal.unuja.ac.id/index.php/core.

[13] A. Budiyantara, I. Irwansyah, E. Prengki, P. A. Pratama, and N. Wiliani, “Komparasi

Algoritma Decision Tree, Naive Bayes Dan K-Nearest Neighbor Untuk Memprediksi
Mahasiswa Lulus Tepat Waktu,” JITK (Jurnal Ilmu Pengetah. dan Teknol. Komputer), vol.
5, no. 2, pp. 265–270, 2020, doi: 10.33480/jitk.v5i2.1214.

[14] M. A. Rahman, N. Hidayat, and A. Afif Supianto, “Komparasi Metode Data Mining K-
Nearest Neighbor Dengan Naïve Bayes Untuk Klasifikasi Kualitas Air Bersih (Studi Kasus
PDAM Tirta Kencana Kabupaten Jombang),” Jurnal Pengembangan Teknologi Informasi
dan Ilmu Komputer (JPTIIK), vol. 2, no. 12. pp. 925–928, 2018.

[15] M. A. Maricar and Dian Pramana, “Perbandingan Akurasi Naïve Bayes dan K-Nearest
Neighbor pada Klasifikasi untuk Meramalkan Status Pekerjaan Alumni ITB STIKOM
Bali,” J. Sist. dan Inform., vol. 14, no. 1, pp. 16–22, 2019, doi: 10.30864/jsi.v14i1.233.

[16] R. P. Fitrianti, “Analisis Sentimen terhadap Review Restoran dengan Teks Bahasa
Indonesia Menggunakan Algoritma K-Nearest Neighbor,” pp. 27–32, 2018.

[17] Yusra, D. Olivita, and Y. Vitriani, “Perbandingan Klasifikasi Tugas Akhir Mahasiswa
Jurusan Teknik Informatika Menggunakan Metode Naïve Bayes Classifier dan K-Nearest
Neighbor,” J. Sains, Teknol. dan Ind., vol. 14, no. 1, pp. 79–85, 2016.

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)
ISSN: 2579-7298 International Journal Of Artificial Intelegence Research
Vol 6, No 1, June 2022

[18] K. U. Syaliman, M. Zulfahmi, and A. A. Nababan, “Perbandingan Rapid Centroid

Estimation (RCE) — K Nearest Neighbor (K-NN) Dengan K Means — K Nearest Neighbor
(K-NN),” InfoTekJar (Jurnal Nas. Inform. dan Teknol. Jaringan), vol. 2, no. 1, pp. 79–89,
2017, doi: 10.30743/infotekjar.v2i1.166.

[19] R. A. Asroni, “Penerapan Metode K-Means Untuk Clustering Mahasiswa Berdasarkan Nilai
Akademik Dengan Weka Interface Studi Kasus Pada Jurusan Teknik Informatika UMM
Magelang,” Ilm. Semesta Tek., vol. 18, no. 1, pp. 76–82, 2015, doi: 10.1038/hdy.2009.180.

[20] W. Dhuhita, “Clustering Menggunakan Metode K-Mean Untuk Menentukan Status Gizi
Balita,” J. Inform. Darmajaya, vol. 15, no. 2, pp. 160–174, 2015.

[21] L. Rusdiana and T. Informatika, “Perbandingan Metode K-Nearest Neighbor Dan Fuzzy C-
Means Dalam Menentukan Predikat Kelulusan Mahasiswa,” vol. 1, no. 1, pp. 21–26, 2017.

[22] K. Kunci, “KOMPARASI METODE KLASIFIKASI DATA MINING UNTUK PREDIKSI

Abstrak,” vol. 5, no. 1, pp. 23–29, 2019.

[23] C. A. Rahardja, T. Juardi, and H. Agung, “Implementasi Algoritma K-Nearest Neighbor

Pada Website Rekomendasi Laptop,” J. Buana Inform., vol. 10, no. 1, p. 75, 2019, doi:
10.24002/jbi.v10i1.1847.

[24] A. Sulistiyo, “ Penentuan Jurusan Sekolah Menengah Atas Menggunakan Metode K-

Nearest Neighbor Classifier Pada SMAN 16 Semarang,” Fasilkom Udinus, vol. 1, no. 1, pp.
1–5, 2014.

[25] W. T. Panjaitan, “Penerapan Algoritma Knn Pada Prediksi Produksi,” J. Univ. AMIKOM
Yogyakarta, pp. 61–66, 2018.

[26] G. A. Pradnyana and A. A. J. Permana, “Sistem Pembagian Kelas Kuliah Mahasiswa

Dengan Metode K-Means Dan K-Nearest Neighbors Untuk Meningkatkan Kualitas
Pembelajaran,” JUTI J. Ilm. Teknol. Inf., vol. 16, no. 1, p. 59, 2018, doi:
10.12962/j24068535.v16i1.a696.

[27] D. Z. Abidin, S. Nurmaini, and R. F. Malik, “Penerapan Metode K-Nearest Neighbor dalam
Memprediksi Masa Studi Mahasiswa ( Studi Kasus : Mahasiswa STIKOM Dinamika
Bangsa ),” Pros. Annu. Res. Semin., vol. 3, no. 1, pp. 133–138, 2017.

[28] H. Risman, D. Nugroho, and Y. Retno, “Penerapan Metode K-Nearest Neighbor Pada
Aplikasi Penentu Penerima Beasiswa Mahasiswa Di Stmik Sinar Nusantara Surakarta,”
TIKomSiN, pp. 19–25, 2012.

[29] Luh Gede Pivin Suwirmayanti, “Penerapan Metode K-Nearest Neighbor Untuk Sistem
Rekomendasi Pemilihan Mobil Implementation of K-Nearest Neighbor Method for Car
Selection Recommendation System,” Techno.COM, vol. 16, no. 2, pp. 120–131, 2017.

[30] Y. Yahya and W. Puspita Hidayanti, “Penerapan Algoritma K-Nearest Neighbor Untuk
Klasifikasi Efektivitas Penjualan Vape (Rokok Elektrik) pada ‘Lombok Vape On,’” Infotek
J. Inform. dan Teknol., vol. 3, no. 2, pp. 104–114, 2020, doi: 10.29408/jit.v3i2.2279.

Guslendra et.al (K-Means and K-NN Methods For Determining Student Interest)