• Dr. P. Mohanaiah
• M. Dileepkumar
• K.Divya
• Sk.Khaja Rasool
• Sk.Hussain Basha

: The rapid growth of social media has created an unprecedented opportunity to detect early-warning signs of mental health crises, including suicidal ideation. Manual monitoring of the billions of posts published daily is wholly impractical, necessitating scalable automated detection systems. This paper proposes a hybrid deep learning framework that synergistically integrates Bidirectional Long Short-Term Memory (BiLSTM) networks, Bidirectional Encoder Representations from Transformers (BERT), and Extreme Gradient Boosting (XGBoost) to identify suicidal intent in Twitter posts. The system is trained and validated on the publicly available "Suicidal Tweet Detection" dataset sourced from Kaggle, comprising 232,074 balanced labeled tweets. BERT provides deep contextual sentence-level embeddings via fine-tuned transformer layers; BiLSTM captures temporal sequential word dependencies; and XGBoost amplifies classification performance through gradient-boosted ensemble learning on a 1,295-dimensional fused feature space. Term Frequency–Inverse Document Frequency (TF-IDF) vectorization and GloVe-100 embeddings serve as complementary feature representations. The integrated model is deployed as a real-time web application using Flask (backend REST API) and HTML/CSS/JavaScript (interactive frontend), enabling binary classification of user-submitted tweets with sub-second inference latency. Experimental results demonstrate a state-of-the-art F1-score of 96.8% and Recall of 97.1%, outperforming all standalone baselines. The proposed system offers a clinically relevant, scalable, and deployable solution to support mental health professionals in crisis detection and early intervention.

Suicidal Ideation Detection; Social Media Mining; BERT; BiLSTM; XGBoost; TF-IDF; NLP; Flask Deployment; Deep Learning; Text Classification; Crisis Intervention; Hybrid Model; Twitter Analysis; Mental Health Monitoring.

"SUICIDAL IDEATION DETECTION ON SOCIAL MEDIA USING A HYBRID DEEP LEARNING FRAMEWORK: LSTM, BERT, AND XGBOOST", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 3, page no.f709-f716, March-2026, Available :http://www.jetir.org/papers/JETIR2603589.pdf

"SUICIDAL IDEATION DETECTION ON SOCIAL MEDIA USING A HYBRID DEEP LEARNING FRAMEWORK: LSTM, BERT, AND XGBOOST", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 3, page no. ppf709-f716, March-2026, Available at : http://www.jetir.org/papers/JETIR2603589.pdf