A Hybrid Attention Based Deep Learning System for Suicidal Ideation Detection in Social Media

Suicide is a critical issue in modern society. Many of the people who have the tendency to suicide share their thoughts and opinions through social media platforms. Suicidal ideation detection via online social network analysis has emerged as an essential research topic with significant difficulties in the fields of natural language processing and psychology in recent years. This paper proposes a hybrid attention-based convolution neural network and long short-term memory (HACNN-LSTM) for suicidal ideation detection of social media data. The proposed system consists of three phases, namely, data preprocessing, word embedding, and classification. To begin, the data preprocessing is carried out on the collected Reddit dataset. After that, the word embedding is performed on the preprocessed dataset by Term Frequency - Inverse Document Frequency (TF-IDF). Finally, the classification is done by using the HACNN-LSTM model, in which the hyperparameter is optimized by particle swarm optimization (PSO) algorithm. The findings proved that the proposed hybrid system achieves superior results than the existing techniques with 96.84% accuracy.