Collaborative Filtering Recommendation Algorithm Based on Cloud Model Clustering of Multi-indicators Item Evaluation

Abstract

Collaborative filtering recommendation algorithm is a personalized recommendation algorithm that is used widely in e-commerce recommendation system. In this paper, a collaborative filtering recomendation algorithm based on cloud model clustering of multi-indicators item evaluation is proposed. In the algorithm, the item evaluation is the object, time weighted function is introduced to item evaluation, soft culsters item based on cloud model and gets the recommended items. The algorithm solves problems of data updating and history validity of evaluation in the collaborative filtering algorithm. Soft cluster item based on cloud model is achieved to avoid the defects bringed by hard division.