Depression tendency detection model for Weibo users based on Bi-LSTM

Abstract

Depression will have a severe impact on social harmony and family happiness. Aiming at users Weibo users, this paper explores the use of deep learning methods. Based on the sentence sentiment analysis task, we propose a depression tendency detection model for Weibo users based on Bi-LSTM. Firstly, Use the Skip-Gram model in Word2Vec to vectorize the text. Adopt Bi-LSTM neural network layer. Through the bidirectional transmission, semantic dependence of capture context, mining the content characteristics of Weibo text. Finally, the text sentiment category is classified through the fully connected layer. The experimental results show that this method can effectively detect the depression tendency for Weibo users.