Proposed Model to Identify Drug Trafficking on Social Media

Abstract

The rise of social media sites, such as Instagram, Telegram, and, has unwittingly provided hidden channels for clandestine sales of illicit drugs, posing daunting challenges to regulatory and law enforcement agencies. Dealers use encrypted messaging, obfuscated language, and temporary or anonymous accounts, which make traditional keyword- based detection methods relatively ineffective. To address these intricacies, the present research presents an advanced Drug Trafficking Detection System that combines machine learning approaches with robust metadata analytics to identify abnormal behaviors and identify dubious activity patterns on social networks. Built on the MERN stack—MongoDB for big data handling, Express.js and Node.js for secure backend services, and React.js for reactive user interfaces the system facilitates real-time data retrieval, abnormality identification, and interactive visualization. Initial assessments suggest that the framework demonstrates high effectiveness in precise identification of drug- related content and interactions. Future work may include the incorporation of deep learning frameworks for intricate content analysis, investigation of privacy- aware surveillance mechanisms, and utilization of federated learning paradigms to enable globally scalable, safe, and efficient surveillance of digital drug trafficking networks.