Deteksi Bunuh Diri pada Media Sosial Twitter Menggunakan Metode CNN-LSTM dengan Ekspansi Fitur Word2vec
Abstrak
Deteksi bunuh diri melalui media sosial telah menjadi tantangan besar dalam beberapa tahun terakhir, terutama pada platform seperti Twitter yang berisi unggahan singkat dan emosional. Penelitian ini bertujuan untuk mengembangkan model deep learning hybrid yang dapat mendeteksi potensi bunuh diri pada tweet di Twitter, menggunakan metode CNN-LSTM dan fitur semantik yang diperluas dengan Word2Vec. Dengan meningkatnya angka bunuh diri di kalangan generasi muda, yang membutuhkan sistem deteksi dini berbasis teknologi canggih. Deteksi dini ini dapat membantu memberikan intervensi lebih cepat bagi individu yang berisiko tinggi. Pendekatan yang diusulkan menggunakan kombinasi Convolutional Neural Network (CNN) untuk menangkap pola lokal dalam teks, Long Short-Term Memory (LSTM) untuk memahami urutan kata dalam teks, serta Word2Vec untuk memperkaya representasi semantik kata-kata dalam tweet. Sistem ini memanfaatkan ekstraksi fitur TF-IDF dan ekspansi fitur menggunakan Word2Vec untuk meningkatkan kemampuan model dalam mengenali pola emosional dan semantik yang ada dalam tweet. Hasil eksperimen menunjukkan bahwa model hybridCNN-LSTM dengan ekspansi fitur Word2Vec dan optimasi menghasilkan akurasi sebesar 91,31%. Hasil model hybrid CNN-LSTM belum mununjukkan hasil yang lebih baik dari model non-hybrid. Kontribusi utama dari penelitian ini adalah mengeksplorasi pengaruh ekspansi fitur Word2vec pada model hybrid deep learning untuk deteksi bunuh diri dan mengintegrasikan ekstraksi fitur TF-IDF sertaoptimasi untuk meningkatkan performa klasifikasi teks.
Kata kunci— deteksi bunuh diri, hybrid deep learning, word2vec, TF-IDF, optimasi
Referensi
World Health Organization, “National Suicide Prevention Strategies: Progress, Examples and Indicators,” Geneva, Switzerland, 2018.
M. Weishaar and A. Beck, “Hopelessness and suicide,” International Review of Psychiatry, vol. 4, Nov. 2009, doi: 10.3109/09540269209066315.
M. A. Silver, “Relation of Depression of Attempted Suicide and Seriousness of Intent,” Arch Gen Psychiatry, vol. 25, no. 6, p. 573, Dec. 1971, doi: 10.1001/archpsyc.1971.01750180093015.
J. Philip, “INTERNATIONAL JOURNAL FOR INNOVATIVE RESEARCH IN MULTIDISCIPLINARY FIELD CNN-LSTM Hybrid Deep Learning Model for Remaining Useful Life Estimation,” 2024, doi: 10.2015/IJIRMF/ICSETI-2024/P04.
T. Mikolov, K. Chen, G. Corrado, and J. Dean, “Efficient Estimation of Word Representations in Vector Space,” Sep. 2013, [Online]. Available: http://arxiv.org/abs/1301.3781
S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Comput, vol. 9, no. 8, pp. 1735–1780, 1997, doi: 10.1162/neco.1997.9.8.1735.
B. R. Babu, S. Ramakrishna, and S. K. Duvvuri, “Advanced Sentiment and Trend Analysis of Twitter Data Using CNN-LSTM and Word2Vec,” in 2025 4th International Conference on Sentiment Analysis and Deep Learning (ICSADL), 2025, pp. 1536–1543. doi: 10.1109/ICSADL65848.2025.10933031.
M. M. Tadesse, H. Lin, B. Xu, and L. Yang, “Detection of Suicide Ideation in Social Media Forums Using Deep Learning,” 2020, doi: https://doi.org/10.3390/a13010007.
H. Kour and M. K. Gupta, “An hybrid deep learning approach for depression prediction from user tweets using feature-rich CNN and bi-directional LSTM,” Multimed Tools Appl, vol. 81, no. 17, pp. 23649–23685, Jul. 2022, doi: 10.1007/s11042-022-12648-y.
X. Rong, “word2vec Parameter Learning Explained,” Nov. 2014, [Online]. Available: http://arxiv.org/abs/1411.2738
Tam CC, Huan N, Cai R, Zhang J, Li z, and Li X, “Evaluation of Artificial Neural Networks in Natural Language Processing to Identify Suicide-Risk Messages on Twitter,” 2022, doi: 10.2196/preprints.42557.
E. Lin, J. Sun, H. Chen, and M. H. Mahoor, “Data Quality Matters: Suicide Intention Detection on Social Media Posts Using a RoBERTa-CNN Model,” 2024. [Online]. Available: http://mohammadmahoor.com
S. Ji, S. Pan, X. Li, E. Cambria, G. Long, and Z. Huang, “Suicidal Ideation Detection: A Review of Machine Learning Methods and Applications,” IEEE Trans Comput Soc Syst, vol. 8, no. 1, pp. 214 – 226, 2021, doi: 10.1109/TCSS.2020.3021467.
T. H. H. Aldhyani, S. N. Alsubari, A. S. Alshebami, H. Alkahtani, and Z. A. T. Ahmed, “Detecting and Analyzing Suicidal Ideation on Social Media Using Deep Learning and Machine Learning Models,” Int J Environ Res Public Health, vol. 19, no. 19, Oct. 2022, doi: 10.3390/ijerph191912635.
Z. Wang and J. Lv, “Data Crawling and Research Based on Topic Web Crawler,” in 2022 International Conference on Computer Network, Electronic and Automation (ICCNEA), 2022, pp. 183–188. doi: 10.1109/ICCNEA57056.2022.00049.
M. E. Weishaar and A. T. Beck, “Hopelessness and suicide,” International Review of Psychiatry, vol. 4, no. 2, pp. 177–184, 1992, doi: 10.3109/09540269209066315.
M. Chatterjee, P. Kumar, P. Samanta, and D. Sarkar, “Suicide ideation detection from online social media: A multi-modal feature based technique,” International Journal of Information Management Data Insights, vol. 2, no. 2, p. 100103, 2022, doi: https://doi.org/10.1016/j.jjimei.2022.100103.
A. K. Uysal and S. Gunal, “The impact of preprocessing on text classification,” Inf Process Manag, vol. 50, no. 1, pp. 104–112, 2014, doi: https://doi.org/10.1016/j.ipm.2013.08.006.
S. Kotsiantis, D. Kanellopoulos, and P. E. Pintelas, “Data Preprocessing for Supervised Learning,” 2014. [Online]. Available: https://www.researchgate.net/publication/228084519
S. Robertson, “Understanding inverse document frequency: On theoretical arguments for IDF,” Journal of Documentation, vol. 60, no. 5, pp. 503–520, 2004, doi: 10.1108/00220410410560582.
E. B. Setiawan, D. H. Widyantoro, and K. Surendro, “Feature expansion using word embedding for tweet topic classification,” in 2016 10th International Conference on Telecommunication Systems Services and Applications (TSSA), 2016, pp. 1–5. doi: 10.1109/TSSA.2016.7871085.
L. Alzubaidi et al., “Review of deep learning: concepts, CNN architectures, challenges, applications, future directions,” J Big Data, vol. 8, no. 1, Dec. 2021, doi: 10.1186/s40537-021-00444-8.
A. Pulver and S. Lyu, “LSTM with Working Memory,” May 2016, [Online]. Available: http://arxiv.org/abs/1605.01988
X. She and D. Zhang, “Text Classification Based on Hybrid CNN-LSTM Hybrid Model,” in 2018 11th International Symposium on Computational Intelligence and Design (ISCID), 2018, pp. 185–189. doi: 10.1109/ISCID.2018.10144.
I. Düntsch and G. Gediga, “Confusion Matrices and Rough Set Data Analysis,” J Phys Conf Ser, vol. 1229, no. 1, p. 12055, May 2019, doi: 10.1088/1742-6596/1229/1/012055.
