Discovering the Common Traits of Cybercrimes in Pakistan Using Associative Classification with Ant Colony Optimization
Abdul Rauf, Muhammad Asif Khan, Hamid Hussain Awan, Waseem Shahzad, Najeeb Ul Husaan · Journal of Cyber Security · 2022
In the modern world, law enforcement authorities are facing challenges due to the advanced technology used by criminals to commit crimes. Criminals follow specific patterns to carry out their crimes, which can be identified using machine learning and swarm intelligence approaches. This article proposes the use of the Ant Colony Optimization algorithm to create an associative classification of crime data, which can reveal potential relationships between different features and crime types. The experiments conducted in this research show that this approach can discover various associations among the features of crime data and the specific patterns that major crime types depend on. This research can be beneficial in discovering the patterns leading to a specific class of crimes, allowing law enforcement agencies to take proactive measures to prevent them. Experimental results demonstrate that ACO-based associative classification model predicted 10 out of 16 crime types with 90% or more accuracy based on discovery of association among dataset features. Hence, the proposed approach is a viable tool for application in forensic and investigation of crimes.