Hoax news-inspector: a real-time prediction of fake news using content resemblance over web search results for authenticating the credibility of news articles

Nowadays social media is one of the important medium of sharing thoughts and opinions of the individual due to its easy access and also it provides an opportunity to the malicious user to post deliberately fabricated false content to influence people for creating controversies, playing with public emotions, etc. The spread of contaminated information such as Rumours, Hoax, Accidental misinformation, etc. over the web is becoming an emergency situation that can have a very harmful impact on society and individuals. In this paper, we have developed an automated system “Hoax-News Inspector” for the detection of fake news that propagates through the web and social media in the form of text. To distinguish fake and real reports on an early basis, we identified prominent features by exploring two sets of attributes that lead to information spread: Article/post-content-based features, Sentiment based features and the mixture of both called as Hybrid features. The proposed algorithm is trained and tested on the self-generated dataset as well as one of the popular existing datasets Liar. It has been found that the proposed algorithm gives the best results using the Random Forest classifier with an accuracy of 95% by considering all sets of features. Detecting and verifying news have many practical applications for business markets, news consumers, and time-sensitive services, which generally help to minimize the spread of false information. Our proposed system Hoax News-Inspector can automatically collect fabricated news data and classify it into binary classes Fake or Real, which later benefits further research for predicting and understanding Fake news.