Prediksi Tingkat Kecanduan Media Sosial pada Remaja Menggunakan Berbagai Algoritma Machine Learning untuk Klasifikasi Multikelas
DOI:
https://doi.org/10.30865/jurikom.v15i5.10139Keywords:
Kecanduan media sosial, machine learning, random forest, support vector machine, decision treeAbstract
Penggunaan media sosial yang berlebihan berkaitan dengan gangguan tidur, kondisi psikologis, aktivitas akademik, dan hubungan sosial pada remaja akhir serta dewasa muda. Penelitian ini bertujuan membandingkan kinerja Decision Tree, Random Forest, dan Support Vector Machine (SVM) linear dalam mengklasifikasikan tingkat kecanduan media sosial menggunakan dataset Students’ Social Media Addiction. Analisis menggunakan 704 observasi setelah satu dari dua baris dengan nilai prediktor yang sama dikeluarkan. Berdasarkan Addicted_Score,data dikelompokkan menjadi kategori low(≤5), medium (6–8), dan high (>8). Student_ID dan Addicted_Score tidak disertakan sebagai prediktor. Data dibagi secara terstratifikasi menjadi 563 observasi latih dan 141 observasi uji. Pengodean kategorikal dan standardisasi fitur untuk SVM dilakukan melalui pipeline, dengan parameter transformasi yang ditentukan hanya dari bagian pelatihan. Pemilihan model didasarkan pada rerata macro F1-score validasi silang lima lipatan. Random Forest memperoleh nilai tertinggi sebesar 0,9708 ± 0,0292, diikuti SVM sebesar 0,9701 ± 0,0165 dan Decision Tree sebesar 0,9334 ± 0,0513. Pada data uji, Random Forest mencapai akurasi 97,87% dan macro F1-score 0,9836, sedangkan Decision Tree menghasilkan macro F1-score tertinggi sebesar 0,9889 dengan akurasi 98,58%. Ketiga model mengenali seluruh 11 observasi kategori high. Random Forest dipilih berdasarkan hasil validasi silang, meskipun selisihnya dengan SVM sangat kecil dan keunggulan statistiknya belum diuji. Hasil penelitian menunjukkan kemampuan model dalam mengklasifikasikan kategori skor pada dataset yang digunakan. Pengujian pada data independen dan verifikasi pembentukan skor diperlukan sebelum model dikembangkan sebagai alat skrining.
References
[1] H. Shannon, K. Bush, P. J. Villeneuve, K. G. C. Hellemans, and S. Guimond, “Problematic Social Media Use in Adolescents and Young Adults: Systematic Review and Meta-analysis,” 2022. doi: 10.2196/33450.
[2] X. Zhu, T. Zheng, L. Ding, X. Zhang, Z. Li, and H. Jiang, “Exploring associations between social media addiction, social media fatigue, fear of missing out and sleep quality among university students: A cross-section study,” PLoS One, vol. 18, no. 10 October, 2023, doi: 10.1371/journal.pone.0292429.
[3] J. Zhuang et al., “A serial mediation model of social media addiction and college students’ academic engagement: the role of sleep quality and fatigue,” BMC Psychiatry, vol. 23, no. 1, 2023, doi: 10.1186/s12888-023-04799-5.
[4] A. Ali Homaid, “Problematic social media use and associated consequences on academic performance decrement during Covid-19,” Addictive Behaviors, vol. 132, 2022, doi: 10.1016/j.addbeh.2022.107370.
[5] B. Hamam, S. Khandaqji, S. Sakr, and A. Ghaddar, “Social media addiction in university students in Lebanon and its effect on student performance,” Journal of American College Health, vol. 72, no. 8, 2024, doi: 10.1080/07448481.2022.2152690.
[6] W. Tafesse, “Social networking sites use and college students’ academic performance: testing for an inverted U-shaped relationship using automated mobile app usage data,” International Journal of Educational Technology in Higher Education, vol. 19, no. 1, 2022, doi: 10.1186/s41239-022-00322-0.
[7] A. Suárez-Perdomo, Z. Ruiz-Delgado, and Y. Garcés-Delgado, “Profiles of undergraduates’ networks addiction: Difference in academic procrastination and performance,” Comput. Educ., vol. 181, 2022, doi: 10.1016/j.compedu.2022.104459.
[8] L. K. Ghanayem, H. Shannon, L. Khodr, R. J. McQuaid, and K. G. C. Hellemans, “Lonely and scrolling during the COVID-19 pandemic: understanding the problematic social media use and mental health link among university students,” Front. Psychiatry, vol. 15, 2024, doi: 10.3389/fpsyt.2024.1247807.
[9] I. Aslan and H. Polat, “Investigating social media addiction and impact of social media addiction, loneliness, depression, life satisfaction and problem-solving skills on academic self-efficacy and academic success among university students,” Front. Public Health, vol. 12, 2024, doi: 10.3389/fpubh.2024.1359691.
[10] C. Hendrikse and M. Limniou, “The Use of Instagram and TikTok in Relation to Problematic Use and Well-Being,” J. Technol. Behav. Sci., vol. 9, no. 4, 2024, doi: 10.1007/s41347-024-00399-6.
[11] H. Lahti et al., “What Counteracts Problematic Social Media Use in Adolescence? A Cross-National Observational Study,” Journal of Adolescent Health, vol. 74, no. 1, 2024, doi: 10.1016/j.jadohealth.2023.07.026.
[12] L. Salerno, L. Fortunato, V. Ostwald, A. Muscolino, and G. Lo Coco, “Social support and social comparison tendencies predict trajectories of adolescents’ problematic social media use: A longitudinal study,” PLoS One, vol. 20, no. 6 June, 2025, doi: 10.1371/journal.pone.0323320.
[13] Y. Hong, X. Rong, and W. Liu, “Construction of influencing factor segmentation and intelligent prediction model of college students’ cell phone addiction model based on machine learning algorithm,” Heliyon, vol. 10, no. 8, 2024, doi: 10.1016/j.heliyon.2024.e29245.
[14] N. T. Thity, A. Rahman, A. Dulmini, N. Yasmin, and R. Rois, “An illustration of multi-class roc analysis for predicting internet addiction among university students,” PLoS One, vol. 20, no. 7 July, 2025, doi: 10.1371/journal.pone.0325855.
[15] D. Zarate, J. Jarrad, V. Stavropoulos, C. Conkey-Morrison, and B. Hunt, “Mental health, personality, and cross-addictions as predictors of social media addiction: a machine learning longitudinal study,” Addictive Behaviors Reports, vol. 24, 2026, doi: 10.1016/j.abrep.2026.100718.
[16] J. R. Quinlan, “Induction of Decision Trees,” Mach. Learn., vol. 1, no. 1, 1986, doi: 10.1023/A:1022643204877.
[17] L. Breiman and A. Cutler, “Breiman and Cutler’s random forests for classification and regression,” Package “randomForest,” 2012.
[18] C. Cortes and V. Vapnik, “Support-Vector Networks,” Mach. Learn., vol. 20, no. 3, 1995, doi: 10.1023/A:1022627411411.
