Suicide Detection from different languages using Python

This is an implementation of the attention-based hybrid architecture (Ghosh et al, 2023) for suicide/depressive social media notes detection.

A pre-trained transfer model based on distilbert-base-uncased for detecting textual suicidal thoughts, utilizing Python

A comparison of three discriminative machine learning models (Logistic Regression, Long Short-term Memory, Transfoermer) to determine which is best suited for suicide ideation detection

Suicide Ideation Detection in Indonesian Language Twitter Posts Using IndoBERT and LSTM/CNN

A deep learning model for detecting suicidal tendencies in text using LSTM and GloVe embeddings. Includes interactive text input with IPyWidgets for real-time analysis.

A RESTful API that embeds classification with confidence of textual suicidal thoughts, conducted on a previously compiled transformer model, utilizing Python

A web application that classifies an entered text into either suicidal or non-suicidal, utilizing Flutter

AI-powered suicide risk classification system with explainable predictions using SHAP and LIME.

A research-focused project aimed at early detection of suicide risk by analyzing both textual and visual data. The system integrates a fine-tuned DistilBERT model combined with CNN for understanding linguistic patterns in user input, and uses YOLO for analyzing facial expressions, gestures, and behavioral cues from images or videos.

Social media suicidality and threat detection designed as a human-on-the-loop system.

🛡️ Classify text messages for suicide risk with a FastAPI service, enabling timely interventions using a 93.33% accurate machine learning model.