Recommendation with Item Clustering Based Collaborative Filtering

Abstract

Recommender systems are playing a more and more important roles in people's daily life and collaborative filtering (short for CF) is a widely used approach in recommender systems.In practice, many E-commerce companies such as Amazon use CF to make recommendations.However, as the number of users and items grow larger and larger, CF are suffering two kinds of problems: sparsity and scalability.So in this paper, we propose an item clustering based CF to solve these two problems.The experiments show that our method outperforms the traditional CF in term of both predicting accuracy and running time.