Research On Movie Recommendation Based On Association Rules Mining

With the growing number of movies,there are more and more films on video websites,which nevertheless often recommend users some low-quality or uninterested movies.How to enhance the quality of recommended movies and movies that can attract user interest have become a hotspot for research of movie recommendation.Despite the gradually complicated movie recommendation system on video websites,as well as the diversity of related recommendation algorithms,the current recommendation system can hardly satisfy user demands.In view of this,to solve the low-quality problem of recommended movies and user interest-related problem,this thesis conducted related research,mainly including the following contents:This thesis firstly introduced the research background and significance,analyzed the effect of applying association rule algorithm to movie labels.Next,this thesis introduced several commonly used movie recommendation algorithms in current movie recommendation system,analyzed their advantages and disadvantages,and proposed several indicators for judging good or poor recommendation algorithms.In the end,this thesis an in-depth analysis of the problems existing in the current movie recommendation algorithm was performed.Secondly,this thesis introduced basic principle of Aprior algorithm is introduced,elaborated the basic thought of weighted Apriori algorithm,and analyzed the implementation process of two algorithms in detail through pseudo-code.Next,this thesis applied the two algorithms to the same example respectively,and analyzed the experiment results,and finally conducted in-depth analysis of the experimental results after comparison between both.Finally,the data set of movies is obtained by crawling the data of popular movies on Douban.Then use the obtained data set to implement a content-based movie recommendation algorithm,and then apply the data to the Apriori algorithm,increase the types of recommended movies,and finally apply the two algorithms to the movie data set,and use the experimental results of the two algorithms to do In contrast,it is verified that the weighted Apriori algorithm can make the movie recommendation effect better,and the recommended movie quality is higher.