Integrasi K-Means dan Decision Tree dalam Menganalisis Kerentanan Pengguna Internet terhadap Kejahatan Siber

Rizki Nabillah, Agustiena Merdekawati, Azis Sukma Dhiana

Abstract


Increasing internet use has heightened exposure to cybercrime, highlighting the need to identify users' vulnerability profiles and distinguishing characteristics. This study integrates K-Means Clustering and Decision Tree C4.5 using questionnaire data from 100 active internet users. K-Means grouped respondents by risk levels and security behaviors into four clusters, selected using the Elbow Method and Davies-Bouldin Index. Cluster labels were subsequently classified using independent predictors, including knowledge, perception, security awareness, age, and gender, to prevent circularity bias. SMOTE was applied exclusively to training data to address class imbalance without data leakage. The model achieved 56.00% accuracy and a macro F1-score of 51.44%. SMOTE increased recall for the low-vulnerability minority class from 30.00% to 40.00%, with 80.00% precision. Cybercrime perception emerged as the primary decision-tree splitting attribute. These findings reveal a knowledge–behavior gap and provide exploratory decision rules to inform behavior-based cybersecurity education.

Meningkatnya penggunaan internet turut memperbesar risiko kejahatan siber sehingga diperlukan pemetaan kerentanan pengguna berdasarkan karakteristik dan perilaku keamanan digital. Penelitian ini bertujuan mengidentifikasi profil kerentanan pengguna internet serta faktor yang membedakan setiap kelompok melalui integrasi K-Means Clustering dan Decision Tree C4.5. Data diperoleh melalui kuesioner terhadap 100 pengguna internet aktif. K-Means digunakan untuk mengelompokkan responden berdasarkan tingkat risiko dan perilaku keamanan, dengan jumlah klaster optimal ditentukan menggunakan Elbow Method dan Davies-Bouldin Index. Selanjutnya, hasil klasterisasi dijadikan label target pada Decision Tree C4.5 dengan memisahkan variabel prediktor berupa pengetahuan, persepsi, kesadaran keamanan, usia, dan jenis kelamin untuk menghindari circularity bias. Ketidakseimbangan kelas ditangani menggunakan SMOTE yang diterapkan hanya pada data latih guna mencegah data leakage. Pengujian menghasilkan akurasi 56,00% dan macro F1-score 51,44%. Penerapan SMOTE meningkatkan recall kelompok kerentanan rendah dari 30,00% menjadi 40,00%, dengan presisi 80,00%. Persepsi terhadap kejahatan siber teridentifikasi sebagai atribut utama dalam pembentukan pohon keputusan. Temuan ini mengungkap kesenjangan antara pengetahuan dan perilaku keamanan (knowledge-behavior gap) serta menghasilkan profil kerentanan dan aturan keputusan yang dapat menjadi dasar pengembangan edukasi keamanan siber berbasis perilaku.

Keywords


Kejahatan siber; Keamanan digital; Perilaku pengguna; K-Means clustering; Decision tree C4.5

Full Text:

PDF

References


Alqahtani, H., & Watson, T. (2024). Human factors in cybersecurity: Evaluating attack surfaces in modern digital environments. Journal of Cybersecurity and Privacy, 4(1), 45-62.

Almufareh, M. F., Al-Ghamdi, A. S., & Mohd, F. (2024). Comprehensive evaluation of clustering validity indices for optimal cluster selection in high-dimensional security data. Journal of Big Data, 11(1), Article 45. https://doi.org/10.1186/s40537-024-00912-x

Amoo, O. O., Adeyemo, A. B., & Omidiora, E. O. (2022). Cybercrime risk assessment using hybrid clustering and machine learning techniques. Computers & Security, 118, Article 102731. https://doi.org/10.1016/j.cose.2022.102731

Ariyaningsih, S., Andrianto, A. A., Kusuma, A. surya, & Rezi. (2023). KORELASI KEJAHATAN SIBER DENGAN PERCEPATANDIGITALISASI DI INDONESIA. Justisia: Jurnal Ilmu Hukum, 1. https://doi.org/10.56457/jjih.v1i1.38

Daeng, Y., Levin, J., Razzaq Prayudha, M., Putri Ramadhani, N., Noverto, Imanuel, S., & Virgio. (2023). Analisis Penerapan Sistem Keamanan Siber Terhadap Kejahatan Siber di Indonesia. INNOVATIVE: Journal of Social Science Research, 3.

Demetz, L., Taylor, S., & Stevens, M. (2023). Rethinking Protection Motivation Theory in digital safety: A contemporary review. Computers & Security, 128, 103150.

Fernández-Navarro, F., Hervás-Martínez, C., & Gutierrez, P. A. (2023). Advanced classification models and resampling techniques for imbalanced data in risk assessment. IEEE Transactions on Cybernetics, 53(4), 2410–2422. https://doi.org/10.1109/TCYB.2022.3189001

Fernández-Navarro, F., Hervás-Martínez, C., & Gutierrez, P. A. (2023). Cost-sensitive learning and class imbalance approaches for cybersecurity risk detection. Information Fusion, 91, 210-224. https://doi.org/10.1016/j.inffus.2023.101804

Han, J., Kamber, M., & Pei, J. (2022). Data mining: concepts and techniques (4th ed.). Morgan Kaufmann. https://doi.org/10.1016/C2013-0-18660-6

ITU. (2026). Hampir tiga perempat populasi terhubung ke internet. ITU. https://www.itu.int/itu-d/reports/statistics/2025/10/15/ff25-internet-use/

Kassambara, A. (2017). Practical guide to cluster analysis in R: Unsupervised machine learning. STHDA.

Krawczyk, B., & Woźniak, M. (2021). Recent advances in imbalanced data classification: Algorithms, applications, and challenges. IEEE Transactions on Knowledge and Data Engineering, 33(8), 2980-2995.

Muningsih, E. (2022). Kombinasi metode K-Means dan Decision Tree dengan perbandingan kriteria dan split data. Jurnal Teknoinfo, 16(1), 112–119. https://doi.org/10.33365/jti.v16i1.1520

Prusty, S. R., Patnaik, S., & Dash, S. K. (2022). Integration of unsupervised clustering and decision tree algorithms for behavioral risk stratification in cybersecurity. Decision Support Systems, 154, Article 113702. https://doi.org/10.1016/j.dss.2021.113702

Sari, S. N., Annisa, P., Rahma, A. N. D., Ritonga, R. P., & Utomo, D. P. (2025). Teknik data mining dengan menggunakan algoritma Decision Tree untuk mengetahui pola pemahaman mahasiswa pada matakuliah pemrograman. Bulletin of Information Technology (BIT), 6(4), 301–309. https://doi.org/10.47065/bit.v6i4.1210

Sarker, I. H. (2023). Cybersecurity data science: Best practices, algorithms, and real-world applications. Journal of Network and Computer Applications, 211, 103560.

Sinaga, K. P., & Yang, M. S. (2020). Unsupervised K-means clustering algorithm. IEEE Access, 8, 80716–80727. https://doi.org/10.1109/ACCESS.2020.2988796

Syafrizal, M., Ashari, M. W., & Cane, M. A. M. (2025). Kerangka Penelitian: Menuju Penyelesaian Studi Dan Penelitian Keamanan Siber. Rabit : Jurnal Teknologi dan Sistem Informasi Univrab, 10(2), 1446–1462. https://doi.org/10.36341/rabit.v10i2.6715

Syakur, M. A., Khotimah, B. K., Rochman, E. M. S., & Satoto, B. D. (2018). Integration K-Means clustering method and Elbow method for identification of the best cluster number. IOP Conference Series: Materials Science and Engineering, 336(1), Article 012017. https://doi.org/10.1088/1757-899X/336/1/012017

Zwilling, M., Kietzmann, J., & Techatassanasoontorn, A. A. (2022). The overconfidence phenomenon in cybersecurity behaviors: An empirical investigation. Computers in Human Behavior, 134, 107312.




DOI: https://doi.org/10.37905/jji.v8i2.41533

Refbacks

  • There are currently no refbacks.