Detection of Depression and Mental illness of Twitter users using Machine Learning

Abstract

Today Micro-blogging has become a popular Internet-user communication tool. Millions of users exchange views on different aspects of their lives. Thus micro blogging websites are a rich source of opinion mining data or Sentiment Analysis (SA) information. Due to the recent emergence of micro blogging, there are a few research works devoted to this subject. We concentrate in our paper on Twitter, one of the prominent micro blogging sites to analyze sentiment of the public. We'll demonstrate, how to gather real-time twitter data for sentiment analysis or opinion mining purposes, and employed algorithms like Term Frequency - Inverse Document Frequency (TF-IDF), Bag of Words (BOW) and Multinomial Naive Bayes ( MNB). We are able to determine positive and negative sentiments for the real-time twitter data using the above chosen algorithms. Experimental evaluations below shows that the algorithms used are efficient and it can be used as a application in detection of the depression of the people. We worked with English in this article, but for any other language it can be used.