Depression Tendency Detection for Microblog Users Based on SVM

With the development of the society, people lay more and more emphasis on mental diseases. Depression accounts for the majority of people all mental diseases. In this paper, a depression detection model based on SVM is proposed to detect whether Sina Weibo (a kind of microblog) users have depression tendency through in-depth mining of Sina Weibo text. First, text features and extended features were extracted. Then SVM model trained with the two kinds of features and fusion features were compared. Through the experiment, the F1 value of the model trained with text features was as high as 84%.