Topic specific emotion detection for retweet prediction

Online social network is a great medium to express one’s opinion, sentiment, preference, and reaction on a topic. Tweets posted by Twitter users are used as a mechanism to share information. By retweeting a tweet, users not only approve the information provided by the tweet but also share the similar emotions and sentiment expressed by the tweet. Analyzing tweets and retweets to discover user’s interest is a challenging and interesting task for researchers mainly in the field of information diffusion. In the past studies, it is usually assumed that a user retweets the tweets which match his topic of interest. However, not only the topic itself but also the emotion and sentiment related to the topic might have impact on user’s retweet decision. In this research, our objective is to explore the impact of user’s topic specific emotion on his retweet decisions. With different latent features, we could find out user’s preferences on different topics at different emotional levels. This research has shown that along with topic, user’s emotion towards a topic is a useful factor in modeling user’s retweet decision.