Software Solutions to Identify Users Behind Social Media Based Drug Trafficking

Abstract

With the fast evolution of social media websites, criminal behavior like drug trafficking can now be performed discreetly using those platforms. Manual detection of such activities becomes challenging owing to huge amounts of unstructured data being generated every single day. This research paper introduces a Drug Trafficking Detection System which automatically analyses tweets for detecting suspicious activities. It utilizes some simple text pre-processing techniques, along with keyword search and risk assessment mechanism, in order to analyze individual tweets. Using a specific risk score generated for each tweet, the tweets are categorized into three types - Low Risk, Medium Risk, and High Risk. This system will generate alerts for risky tweets that police can act upon immediately. The system minimizes human labor, increases accuracy, and makes detection quick. The system is easy to use, inexpensive, and scalable; therefore, it is applicable in real-life settings. Despite its reliance on the current model where detection is based on keywords, this system can be improved through time to include sophisticated machine learning models and real-time data analysis. All in all, this system serves as an efficient method of monitoring social media to fight drug abuse