A Model for Improving Recommender Systems Based on the Similarity of Users

Todays, regarding the information overload on the web, providing services to enhance network efficiency by using computer-based recommender systems has become a necessity. Naturally, these systems are unable to provide recommendations without the appropriate information. Therefore, the accuracy of the recommendation in the recommender systems is one of their main challenges. Hence, the purpose of this study is to provide a model for increasing the accuracy of recommender systems using two commonly used methods in recommendatory systems (clustering algorithm and associative rules)based on the possible similarity of users using Pearson Correlation Coefficient and the Crisp method. The data used to carry out this research is taken from the MovieLens site. The K-means algorithm was used for clustering, and the Apriori algorithm was used to extract rules for films and to determine the degree of films' dependency. The results indicated that the proposed model was more accurate than previous models.