Implementation of Artificial Intelligence to Predict Threats in Social Media Based on User’s Behavior

Abstract

The online social network users are with faced multiple vulnerabilities threats that are strategically implementing the social engineering mechanisms. Cybercriminals targeting the social engineering technique often investigate the environment in which a user is subjected. However, current research is pegged on the technical measurement of how threats can be neutralized or prevented completely in the online social network environment. It, therefore, follows that the online social networks systems create characteristics and elements which can be investigated further to produce relevant models. Facebook in the recent past has worked so hard and invested heavily in developing algorithms that are able to determinate an imminent cyber-attack based on user's behavior and characteristics on the platform. As much as Facebook has done relevant technical measures to ensure that threats are minimized as much as possible there is still a gap of further investigations which seek to harness the automated algorithmic prediction using artificial intelligence to determine the possibility of an attack or a threat. This research uses the machine learning techniques to show how an AI-based algorithm of the user's behavioral characteristics, perceptions, and socio-emotions that would help in identifying features that further become very relevant in determining an individual's vulnerability to social engineering threats and attacks. The aim is to primarily study the habitual perspective and perpetual perspective reactions towards socially engineered threats daily. This research uses a comparative analysis of the theories and the findings from the primary data to show that certain behaviors of Facebook users are a threat to other users. Through machine learning techniques, the results have shown how A.I. algorithms work in identifying scamming messages and scammers.