Recommendation System for Criminal Behavioral Analysis on Social Network using Genetic Weighted K-Means Clustering

Abstract

The accessibility and usage of social networking sites constructs both prospects and menaces for the users. In this research article, we propose a new recommendation system for predicting and recommending the criminal behavioral users on social network based upon the activities of the users. Our recommender system uses the proposed nine factor analysis method, clustering technique called Genetic Weighted K-Means clustering (GWKMC) and the existing classification algorithm namely Negative Selection Algorithm (NSA). The proposed recommendation system is evaluated by conducting various experiments using the Face book dataset (Latest) which is prepared on our own and also the Weblog dataset (Timeworn). The conducted experiments confirmed the efficacy of the proposed Recommender System.