Sentiment and Emotion Analysis From Psychological Support Groups of Facebook

Abstract

Sentiment and emotion analysis is the way of discovering positive, negative and neutral sentiment in text. In our paper we try to analyze the sentiment from posts/comments in psychological support groups in the most popular social media platform Facebook. The advantage of social media is it’s not only used for entertainment but also for helping people who need help. Nowadays many people suffer from depression, it's one of the most serious mental health problems. psychological support groups on Facebook try to help those depressed people. We analyze many posts/comments on psychological support groups on Facebook and predict how much those posts/comments can support or help a person. In our project we collect practical examples that are posted in different psychological groups, and we calculate those posts as helping people to improve their situation. We can know the effect of how supportive these posts are or not and how this post is helping people. Is it in position, positive or neutral way? For this task we divided the complete work into two sections: sentiment detection and analyzing the ability to detect sentiment from such a special category of texts. For visualization here we use Matplotlib, Seaborn, NumPy. For graph visualization we use scatterplot, word cloud and for visualization we bring word cloud from monkey learning website. For overall tasks we have utilized Natural Language Toolkit (NLTK) and TextBlob, which are publicly available python packages.