Collaborative filtering for music recommender system

Nowadays recommender systems is a software that is used as an important tool of e-commerce, which helps to analyze users' tastes and provide them with lists of products that they would like to prefer. This paper is an investigation of using collaborative filtering techniques for a music recommender system. Collaborative filtering is the technology that focuses on the relationships between users and between items to make a prediction. The goal of the recommender system is to compute a scoring function that aggregates the result of computing similarities between users and between items. We focus on the reviewing two strategies of collaborative filtering: user-based and item-based recommendations. For experimental purpose we explore different metrics to measure the similarity of users and items such as Euclidean distance, cosine metric, Pearson correlation and others. Finally, we compare different evaluations metrics that represent the effectiveness of the recommender system.