Machine Learning and User Social Engagement based approaches for Fake News Fake Accounts detection in OSN

Abstract

Social Network is a collection of social actors or firms that connect and share information. The utilization of Online Social Networks (OSN) as the primary source of information has increased, along with the availability of online social platforms that provide news from various sources. The trend of obtaining information from diverse sources also led to the scope of spreading fake news. Fake News or Yellow Journalism (Campbell, 2019) can be an intentional or unintentional spread of unreal or unverified information. Fake News propagation is regarded as one of the main problems of online social networks, whose spread relies on social influence and psychological theories (like Social Credibility, Bias, Echo Chamber Effect, etc.). As huge volumes of information are being published from various sources, it is hard to check the credibility of the information users consume. After post-truth (user s tendency to accept an argument based on emotions and beliefs rather than facts) was named the Word-of-the-year by Oxford University in 2016, there have been more discussions about Fake News. Even though many automatic fact-checking sites are available online (like Thip.media, The Quint, etc.), there is a huge need to develop promising models to detect fake news and accounts in OSNs. More sophisticated tools are required to analyze the online social media being populated. An online social network refers to the connections and relationships between individuals, and online social media refers to a broader category that encompasses various web-based platforms and applications that enable users to create, share, and interact with content. Social media platforms can include Online Social Networks as well as other forms of user-generated content, such as blogs, microblogs, photo-sharing sites, video-sharing sites, and forums (Knoke and Yang, 2019). Data analytics tools are being used to find the insights from this voluminous data and make necessary decisions.