Enhancing Long Tail Recommendation Based on User's Experience Evolution

At present, recommender system has become a hot research area. It has been widely used, especially in e-commerce, news promotion, online education and other fields. Collaborative filtering algorithm based on user behavior analysis is a popular algorithm in recommender system research. In the research of user taste evolution, most of the work considered from the perspective of the group, thus ignoring the difference of individual learning ability of users. On the other hand, how to effectively mine long tail items has become one of the problems that need to be solved in recommender systems. In this paper, we model the evolution of user tastes from the individual level by user rating sequence. Then we build the recommendation scoring model according to the user experience level and the corresponding popularity of the items, and make reasonable topN recommendation. Experiments show that our model can achieve good results. In addition, the recommendation based on user experience level can effectively solve the long tail mining problem.