Collaborative Filtering Recommendation Algorithm Based on Improved BiasSVD

In the era of information explosion, the internet platform with massive information provides personalized services for users by using recommendation system. With users and projects growing ever more, the data sparsity problem may affect the cold start of projects. To solve the above problem, an improved collaborative filtering recommendation algorithm is proposed in this paper. Firstly, the biasedSVD algorithm is used to reduced dimension of user rating matrix, which brings the less complexity and alleviates the problem of data of sparsity. Secondly, the user's behaviors are analyzed and a novel weight named the time-preference weighting factor is proposed to describe the user's rating preference for items. Experimental results based on MovieLens data set show that the proposed method outperforms the state-of-the-atrs.