Crime Intention Detection Using Ontology in Social Media Platforms

Today’s social media platforms (SMPs) serve as a core for criminal activity. Law enforcement organizations are trying to filter and analyze the massive amounts of data from such platforms. The internet and online internet tools are crucial in bringing people together who have similar interests. Unfortunately, such intimate contact makes it possible for illegal activities to take place. The SMPs are used by criminals for a variety of criminal purposes, including the creation of criminal virtual groups, user information sharing, and data breaches. None of the ontology paradigms among those discovered in the research enable a complete evaluation of all elements of criminal information sharing, data breaches, and unwanted crimes in SMPs. Hence, we suggested an ontology-based gradient descent optimized AdaBoost algorithm (OGDOAA) for criminal intention detection in SMPs, which creates program concepts for the choice of social network postings containing criminal slang terms and automatically categorizing these posts in line with illocutionary categories. The system uses trained models from previously published articles to accurately classify published posts with criminal purposes. The suggested method is examined and contrasted with different existing technologies. The results show that the suggested framework is effective in identifying crimes on social media.