Deteksi Cyberbullying Tiga Kelas pada Komentar Berbahasa Indonesia di Aplikasi X dan Threads Menggunakan IndoBERTweet

Nisya Indriyani; Chanifah Indah Ratnasari

doi:10.56427/jcbd.v5i2.1009

Authors

Nisya Indriyani Universitas Islam Indonesia
Chanifah Indah Ratnasari Universitas Islam Indonesia https://orcid.org/0000-0001-7328-8242

DOI:

https://doi.org/10.56427/jcbd.v5i2.1009

Keywords:

Cyberbullying, IndoBERTweet, Media sosial, Klasifikasi teks, Ambiguitas

Abstract

Media sosial menjadi ruang utama untuk berinteraksi, tetapi juga memunculkan risiko cyberbullying melalui komentar yang menghina, merendahkan, atau menyerang pengguna lain. Penelitian ini bertujuan mendeteksi cyberbullying pada komentar berbahasa Indonesia dari aplikasi X dan Threads menggunakan IndoBERTweet. Penelitian menggunakan pendekatan kuantitatif dengan desain supervised machine learning. Dataset akhir terdiri atas 9.182 komentar yang diseimbangkan berdasarkan platform, yaitu 4.591 komentar dari aplikasi X dan 4.591 komentar dari Threads. Sebagai kebaruan, penelitian ini menerapkan klasifikasi tiga kelas, yaitu non-cyberbullying, ambiguitas, dan cyberbullying, untuk mengakomodasi komentar yang belum dapat dikategorikan secara tegas sebagai aman atau berbahaya. Proses penelitian meliputi pengumpulan data berbasis kata kunci, pelabelan awal berbasis aturan, validasi pakar pada data terpilih, prapemrosesan terbatas, fine-tuning IndoBERTweet, evaluasi model, dan integrasi ke dalam prototipe sistem. Hasil pengujian pada 919 data uji menunjukkan accuracy sebesar 0.7911 dan macro F1-score sebesar 0.7294. Model memperoleh F1-score sebesar 0.8680 pada kelas non-cyberbullying, 0.4901 pada ambiguitas, dan 0.8301 pada cyberbullying. Evaluasi platform menunjukkan accuracy sebesar 0.8214 pada aplikasi X dan 0.7609 pada Threads. Model dapat digunakan sebagai alat bantu moderasi awal, tetapi komentar ambigu tetap memerlukan peninjauan manusia.

Downloads

Download data is not yet available.

References

[1] S. Suhendra and F. Selly Pratiwi, “Peran Komunikasi Digital dalam Pembentukan Opini Publik: Studi Kasus Media Sosial,” IAPA Proceedings Conference, pp. 293–315, Oct. 2024, doi: 10.30589/proceedings.2024.1059.

[2] G. Alexandro, J. Periantalo, and M. A. Rahman, “Hubungan Regulasi Emosi dan Perilaku Cyberbullying pada Mahasiswa Dewasa Awal Pengguna Instagram dan TikTok di Kota Jambi,” Observasi : Jurnal Publikasi Ilmu Psikologi, vol. 4, no. 1, pp. 241–249, Feb. 2026, doi: 10.61132/observasi.v4i1.2126.

[3] P. Wilujeng, I. Muslim, and Sunariyo, “Analisis Yuridis Tindak Pidana Cyberbullying Berdasarkan pada Undang-Undang ITE dan Kitab Undang-Undang Hukum Pidana,” Jurnal Hukum Lex Generalis, vol. 5, no. 12, pp. 1–12, Feb. 2024, doi: 10.56370/jhlg.v5i12.827.

[4] E. Asalnaije, Y. Bete, M. A. Manikin, R. A. Labu, S. A. D. Tira, and Y. P. Lian, “Bentuk-Bentuk Cyberbullying Di Indonesia,” INNOVATIVE: Journal Of Social Science Research, vol. 4, no. 4, pp. 6465–6473, Jul. 2024, doi: 10.31004/innovative.v4i4.12471.

[5] A. W. Utami and I. D. Arianto, “Perilaku Cyberbullying pada Media Sosial TikTok: Analisis Isi Kualitatif Perilaku Cyberbullying di Kolom Komentar dalam Akun TikTok (@ofp24),” Linimasa: Jurnal Ilmu Komunikasi, vol. 7, no. 2, pp. 1–20, Aug. 2024, doi: 10.23969/linimasa.v7i2.11484.

[6] Iswandi, S. A. Nasution, E. Citrowati, Y. D. Septiana, and Salman, “Cyberbullying pada Anak Sekolah Dasar: Studi Fenomena di Era Digital,” Pendas: Jurnal Ilmiah Pendidikan Dasar, vol. 11, no. 1, pp. 191–207, Mar. 2026, doi: 10.23969/jp.v11i01.42093.

[7] A. R. Daulay, F. L. D. Prayoga, F. Hutagalung, I. N. Y. Marbun, N. A. Primadiati, and S. W. Simbolon, “Pengaruh Nilai Pancasila pada Cyberbullying di Media Sosial (Kasus Kekerasan Verbal oleh Luluk Sofiatul Jannah),” NuSCJo: Nusantara Community Service Journal, vol. 2, no. 2, pp. 147–152, Dec. 2025, doi: 10.70437/nucsjo.v2i2.288.

[8] R. R. Aditya and A. M. Syarif, “Deteksi Cyberbullying pada Komentar Media Sosial Berbahasa Indonesia Menggunakan Pendekatan Hibrida IndoBERTweet-BiLSTM,” Building of Informatics, Technology and Science (BITS), vol. 7, no. 4, pp. 2322–2334, Mar. 2026, doi: 10.47065/bits.v7i4.9249.

[9] M. V. S. Handayani and Muljono, “Arsitektur Hibrida IndoBERTweet-CNN untuk Klasifikasi Ujaran Kebencian Berbahasa Gaul di Media Sosial,” Infotekmesin, vol. 17, no. 1, pp. 39–47, Jan. 2026, doi: 10.35970/infotekmesin.v17i1.3026.

[10] E. W. Pamungkas, V. Basile, and V. Patti, “Investigating the role of swear words in abusive language detection tasks,” Language Resources and Evaluation, vol. 57, no. 1, pp. 155–188, Mar. 2023, doi: 10.1007/s10579-022-09582-8.

[11] P. Chiril, E. W. Pamungkas, F. Benamara, V. Moriceau, and V. Patti, “Emotionally Informed Hate Speech Detection: A Multi-target Perspective,” Cognitive Computation, vol. 14, no. 1, pp. 322–352, Jan. 2022, doi: 10.1007/s12559-021-09862-5.

[12] M. F. Ruziqiana, L. Hidayah, and M. A. Rasyidi, “Detection of Cyberbullying Using SVM, Naive Bayes, and Random Forest Algorithm,” Jurnal Informatika dan Teknik Elektro Terapan, vol. 12, no. 3, pp. 3384–3392, 2024, doi: 10.23960/jitet.v12i3.5283.

[13] K. Lukman and S. Novianto, “Komparasi Algoritma Naïve Bayes dan SVM untuk Identifikasi Cyberbullying Selebriti di Media Sosial Twitter,” Jurnal Algoritma, vol. 22, no. 1, pp. 970–981, Jun. 2025, doi: 10.33364/algoritma/v.22-1.2196.

[14] Y. Sagama and A. Alamsyah, “Multi-Label Classification of Indonesian Online Toxicity using BERT and RoBERTa,” 2023 IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT), pp. 143–149, Jul. 2023, doi: 10.1109/IAICT59002.2023.10205892.

[15] U. Khairani, V. Mutiawani, and H. Ahmadian, “Pengaruh Tahapan Preprocessing Terhadap Model IndoBERT Dan IndoBERTweet Untuk Mendeteksi Emosi Pada Komentar Akun Berita Instagram,” Jurnal Teknologi Informasi dan Ilmu Komputer, vol. 11, no. 4, pp. 887–894, Aug. 2024, doi: 10.25126/jtiik.1148315.

[16] J. Haber et al., “Improving the Detection of Multilingual Online Attacks with Rich Social Media Data from Singapore,” Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics, vol. 1, pp. 12705–12721, Jul. 2023, doi: 10.18653/v1/2023.acl-long.711.

[17] A. C. Nurcahyo, H. Y. Ting, and A. F. Atanda, “Network Log Implementation for GRU Based Bandwidth Classification,” Journal of Computers and Digital Business, vol. 4, no. 2, pp. 76–89, May 2025, doi: 10.56427/jcbd.v4i2.763.

[18] A. Vazhentsev et al., “Hybrid Uncertainty Quantification for Selective Text Classification in Ambiguous Tasks,” Association for Computational Linguistics, vol. 1, pp. 11659–11681, Jul. 2023, doi: 10.18653/v1/2023.acl-long.652.

[19] M. Sandri, E. Leonardelli, S. Tonelli, and E. Jezek, “Why Don’t You Do It Right? Analysing Annotators’ Disagreement in Subjective Tasks,” Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pp. 2428–2441, May 2023, doi: 10.18653/v1/2023.eacl-main.178.

[20] M. Taulé, M. Nofre, V. Bargiela, and X. Bonet, “NewsCom-TOX: a corpus of comments on news articles annotated for toxicity in Spanish,” Language Resources and Evaluation, vol. 58, no. 4, pp. 1115–1155, Dec. 2024, doi: 10.1007/s10579-023-09711-x.

[21] C. H. Lee, H. C. Yang, Y. J. Chen, and Y. L. Chuang, “Event monitoring and intelligence gathering using twitter based real-time event summarization and pre-trained model techniques,” Applied Sciences (Switzerland), vol. 11, no. 22, Nov. 2021, doi: 10.3390/app112210596.

[22] P. Zhang, Y. He, E.-U. Haq, J. He, and G. Tyson, “The Emergence of Threads: The Birth of a New Social Network,” arXiv preprint, Jun. 2024, doi: 10.48550/arXiv.2406.19277.

[23] M. L. Williams, P. Burnap, and L. Sloan, “Towards an Ethical Framework for Publishing Twitter Data in Social Research: Taking into Account Users’ Views, Online Context and Algorithmic Estimation,” Sociology, vol. 51, no. 6, pp. 1149–1168, Dec. 2017, doi: 10.1177/0038038517708140.

[24] J. Sadaiyandi, P. Arumugam, A. K. Sangaiah, and C. Zhang, “Stratified Sampling-Based Deep Learning Approach to Increase Prediction Accuracy of Unbalanced Dataset,” Electronics (Switzerland), vol. 12, no. 21, Nov. 2023, doi: 10.3390/electronics12214423.

[25] F. Koto, J. H. Lau, and T. Baldwin, “IndoBERTweet: A Pretrained Language Model for Indonesian Twitter with Effective Domain-Specific Vocabulary Initialization,” in Association for Computational Linguistics, Sep. 2021. doi: 10.18653/v1/2021.emnlp-main.833.