A Personalized Recommendation System Based on User Portrait

Abstract

This paper focus on video recommendation in response to the explosive growth of movies which has made people in the dilemma of information overload for a long time. It is particularly important to push movies accurately. Regarding this situation, this paper studies how to recommend the most similar movies to users based on their interest tags and browsing behavior. Compared with the collaborative filtering algorithm, whose processing effect is not good confronting scoring data sparse problem. This paper makes full use of the movie review data, and the experimental data set is obtained by crawling the high-quality review of the Douban website. The experimental results show that the proposed method can effectively solve the problem of difficulty in calculating the similarity of the traditional recommendation algorithm when the user data set is incomplete, and the recommended effect is satisfactory.