Predicting the Veracity of Fake Information on Smart Media for Reducing Misinformation Diffusion

Abstract

Now is “The digital era”. In this era the data or information plays a very critical role. In this phase of human history that may traits of day to day activities in life have undergone drastic change in ways of trade and commerce, communication medium, health care, media ecosystem and many other aspects. One such shift is observed in consumption of information. Majority of adults of Smart City consume information/news on Smart Media i.e. Social Media platforms like WhatsApp, Facebook, Twitter, YouTube and misc. This dependency further accounts vulnerability to those as a large mass is being manipulated by the spread misinformation. Making this scenario more challenging is the advent of Big Data disposal in form of articles, headlines, videos, tweets, posts and hashtags. Due to variety of sources of information, the acceptability of the information has become a severe apprehension. Recently this issue has grabbed attention of researchers and various algorithms have been devised and methodologies have been adapted to address this issue as it can not only affect political but also social aspect. This paper focuses on literature study of various approaches adapted to combat the issue of veracity of misinformation on smart media for reducing misinformation diffusion. Analysis of various Machine Learning (ML) algorithms for classification of veracity of tweets is also made. A generic model to combat this issue is derived. Further the scope of research work in this area has also been discussed. The dataset for experimentation is trained and evaluated on twitter data consisting of rumours collected from several real-world events.