Proposed Model to Find Most Prominent Node in Social Media Network for Cyber Crime Detection Using Modified Cluster Walktrap and Analytical Hierarchy Process

Abstract

With the proliferation of extensive network structures and real-time information exchange, social media platforms have become pivotal sources for the detection of criminal activities. This study presents a two-phase computational framework designed to identify and analyze criminal communities within social media networks. In the initial phase, a modified Cluster Walktrap algorithm is employed to detect criminal communities by leveraging the graph-theoretic properties of social media interactions. This approach enhances the accuracy of community detection by optimizing the clustering process to address the distinctive characteristics of criminal networks. In the subsequent phase, the Analytical Hierarchy Process (AHP) is utilized to identify the most influential node within the detected criminal communities. AHP integrates multiple parameters, including degree centrality, betweenness centrality, and mean distance, to systematically evaluate and rank nodes based on their influence within the network. The incorporation of multiple parameters enhances the robustness and reliability of the results. The proposed framework represents a significant advancement in criminal network analysis by integrating sophisticated clustering methodologies with multi-criteria decision-making techniques, thereby improving the accuracy and interpretability of criminal activity detection in social media environments.