Fake Review Detection in Movie Recommendation System Using Hybrid Method

Abstract

The paper proposes a hybrid method for detecting fake reviews in movie recommendation systems to eliminate the negative effects of artificial reviews on user choices. The proposed system uses a combination of network and profile features of the user to identify accounts posting fake reviews to promote or demote a particular movie. The network features include location and network usage while the profile features include account activity and creation date. The IP address of the fake accounts is then blocked to prevent them from further posting fake reviews. Real reviews from genuine accounts are used for movie recommendations, providing users with accurate and reliable recommendations based on their preferences. The objective is to enhance user trust in the recommendation system and to provide users with more reliable and authentic reviews. The study shows that the proposed hybrid method outperforms existing methods for fake review detection in terms of precision, recall, and accuracy. The results indicate that the proposed method significantly improves the credibility and reliability of recommendation systems by removing misleading and fake reviews.