Analyzing the social conduct of users to detect and prevent cybercrime across social networks

The online social network features have made it possible to reach cybercriminals in previously unreachable countries and locations. Machine Learning techniques allow us to predict and detect harmful forms of human behaviour such as cyberbullying. Cyberbullying can be easily committed; it is fast-spreading and dangerous, aggressive behaviour. The criminal only requires a laptop and an internet connection and no confrontation with the victim. Anonymous users on social media websites are also the primary cause of aggressive behaviour. Big-data analysis unfolds the otherwise hidden information through deep learning and can help forecast the future when combined with machine learning techniques. The rise of cybercrime poses a significant threat to individual users and the community. For that, effective mitigation techniques are urgently required. This study proposes an innovative model IdentityLinkerNet (ILN) designed to detect cybercriminals by analyzing the social conduct of users across social networks. There are some ethical concerns regarding this study which are privacy concerns related to data collection and sharing, consent and data security. The limitation of this study lies in collecting datasets for user behaviour analysis across multiple social media platforms.