Identification Of Internet Users Mental Health Based On Sina Micro-blog

Currently, the problem of mental health(depression) has become an important issue besides body health. This issue not only has a great deal of harm for the individual, but it has a big impact for the whole social environment.In view of the weakness of present traditional psychology measurement in mental health, this paper presents a mental health(depression) recognition method through use social media behavior of Internet users. This method is to use machine learning to build a prediction model of depression.Firstly, this paper analyzes the present research situation and advantages/disadvantages of the Internet social media behavior and mental health status. Secondly, we discusses the feasibility of using social media behavior to predict depression. In view of some disadvantages of scale measurement, we present a process to alternate traditional method and implement the depression prediction model based on users’ social media behaviour. Then based on the different periods of original data, we explore the method to improve the accuracy of the prediction model. In the end, we build a mental health display system and prospect the mental health aid work.In the example of this paper, we build a prediction model based on linguistic and behavior features of 10,102 Sina Weibo users. By utilizing classification technique, we explored the best prediction outcome across multiple time periods. Results indicated that users’ depression can be detected via Sina Weibo data.Besides, if we want to optimize the performance of depression prediction in current time, we need to observe users’ data before half a month and collect data with a length of two months.