Similarity Based Collaborative Filtering Model for Movie Recommendation Systems

Recommendation systems are information filtering tools and techniques which provide valuable suggestions to users about items they may like by considering various attributes. In Recommendation algorithms, similarity metrics are the main components and performances of these will have direct influence on the recommendations generated. Selecting suitable Similarity measures will be a major factor in improving performance of data mining techniques. These are mainly used for finding the similarity between the users or items as per the requirement for different purposes. In this paper, performance of various similarity metrics applied in recommendation tasks are compared by considering MovieLens dataset for developing a recommendation system for movies using user-based and item-based.