Taxonomy of Shilling Attack Detection Techniques in Recommender System

Abstract

Collaborative filtering recommender systems (CFRSs) are widely employed in several applications because of their satisfying performance in the customized recommendation. Recent studies show that CFRSs are at risk of shilling attacks due to injection of some shilling profiles by the associate resister into a system. The genuineness of the rated recommendations as well as users’ trustworthiness are getting seriously affected due to the injection of malicious users within the recommendation system. Over the last few decades, numerous studies have been conducted to survey shilling attack detection schemes and study recommendation algorithms and different profile injection attack techniques. In this chapter, the profile injection attack methods, collaborative filtering in recommender systems and shilling attack detection schemes have been discussed. The shilling attack detection technique basically works on the two parameters, namely, rating parameter and rating and time interval parameter. Based on the working parameters, the shilling attack detection techniques have been classified. They have also been classified based on their output. A brief discussion on each and every scheme has also been presented. The techniques have been analyzed based on their advantages and disadvantages.