Analysis of Information Pollution on Social Media Platform

Social networking sites have made the communication very easy and each one can express its views and opinions. But this also opens up an opportunity of spreading anecdotal information intentionally. With the increase in content on social network, it is not easy to differentiate between fake information or genuine information. In this paper, authors have analyzed the anecdotal information based on different social platforms and proposed a methodology for the detection of fake news using Term Frequency & Inverse Document frequency (TFIDF) feature vector and different classification models. Authors have analyzed that random forest classifier outperformed best with accuracy upto 95.75%.