Implicit Recommendation with Interest Change and User Influence

Abstract

Aiming at the problem of rich websites in campus without targeted recommendation, which makes it difficult for users to find the information resources of high interest and high quality, this paper proposes an implicit feedback recommendation algorithm in campus network based on user's changing interest and user influence. Based on the traditional collaborative filtering algorithm, introduces time function that adapting to user's changing interest and user's influence factors. The score matrix based on time weight is integrated with the influence matrix to solve the problem that user similarity calculation is too single, and improves the accuracy and explanatory of the recommendation results. Experimental results show that the algorithm can effectively reduce the sparsity and cold start problem of the dataset, and has better recommendation quality than traditional collaborative filtering algorithm.