Contextual Semantic analysis on Blogs/Websites and its Credibility

Abstract

As technology advances, there is a great deal of deceptive information revolving around the news websites and all-over the social media. This menace can only be handled by automating the verification of authenticity of the websites rather than manually detecting them. This research presents a novel method for the verification of the credibility of the news. This method includes web-crawling, sentence scoring, and comparison with authentic websites. Initially, keywords for a topic from twelve verified websites are used for training the Relational Description Framework (RDF) based database. Then, a contextual semantic analysis of the news from a blog/website is done, and keywords are identified. These keywords are compared with the trained keywords of the RDF based database using Term Frequency-Inverse Document Frequency (TF-IDF) and Logistic Regression algorithms. The highest accuracy achieved is 82.8% after a blog/website content is compared with the database. This will help to identify only legitimate news.