Opinion Mining from Twitter Data using Probabilistic Logic Reasoning
The explosion of user-generated content in social media has generated a shift on the way people interact on the web and influence each other. Twitter is one of the most active social networks, with millions of tweets sent daily, where multiple users express their opinion about travelling, economic issues, political decisions etc. This article focuses on opinion mining in Twitter data, using probabilistic logic reasoning. Our approach uses a Bayesian-based opinion mining framework exploiting Twitter Data. The framework of our approach imports Tweets massively by using Twitter’s API. The imported Tweets are further processed automatically for constructing a set of untrained rules and random variables. An evidence set and the sets of untrained rules and random variables are used for the derivation of a Bayesian Network. The derived trained model can then be used for the evaluation of new Tweets, whereas a novelty of our approach is that the constructed model can be incrementally retrained thus becoming more robust. As an application domain, we have selected tourism as it is one of the most popular topics in social media, effectively predicting users’ intention to visit a place. Additional advantages of our system are its ability to be easily adapted to opinion mining from social media on topics other than tourism. Finally, the rules of the derived model are constructed automatically in an efficient way.
Bayesian Networks; Twitter; Statistical Relational Learning; Parameter Learning.