Journal of Jilin University Science Edition

Previous Articles     Next Articles

Personalized Recommendation Method of Film Based on Reviews

DONG Xueyang1, GAO Shanquan2, LIU Lei2, LIU Huaxiao3   

  1. 1. Public Computer Education and Research Center, Jilin University, Changchun 130012, China;2. College of Software, Jilin University, Changchun 130012, China;3. College of Computer Science and Technology, Jilin University, Changchun 130012, China
  • Received:2017-06-13 Online:2018-07-26 Published:2018-07-31
  • Contact: DONG Xueyang E-mail:dxy1986121@jlu.edu.cn

Abstract: Aiming at the problem that there were a large number of films, and it was difficult for audiences to choose their favorite films, we proposed a personalized recommendation method of film based on reviews. Firstly, the method classified the audiences’ reviews by using the film labels as the topics of film. Secondly, we calculated the satisfaction degree of topics according to the evolution function of the topic. Thirdly, we set the weights of topics according to audiences’ preference and obtained the personalized evaluation of the film. Finally, according to the evaluation results, audiences were provided with an alternative sequence of films. The experimental results show that the analysis method of film based on reviews is effective.

Key words: personalize, analysis of review, recommendation method, film review

CLC Number: 

  •