Font Size: a A A

Research On Movie Recommendation System Based On User Interest And Relationship

Posted on:2022-06-05Degree:MasterType:Thesis
Country:ChinaCandidate:L J WangFull Text:PDF
GTID:2505306347956049Subject:Master of Engineering
Abstract/Summary:PDF Full Text Request
With the development of Internet technology,data in various fields has exploded,and the problem of data overload is very prominent.It takes a lot of time for users to obtain the data they are interested in.To solve this problem,recommendation technology came into being to provide users with information that meets their needs,and has been widely used in the field of film and television.However,with the in-depth development of movie recommendation,there have been problems of solidification of recommendation results and inaccurate recommendation of new users.In response to the above problems,this paper designs a recommendation model based on user interests and a recommendation model based on user relationships,and combines the two models to design and develop a movie recommendation application system based on user interests and relationships.The main work is as follows:First of all,traditional movie recommendation has the problem of solidification of recommendation results,which will cause users to suffer from visual fatigue and reduce the sense of experience.In this paper,long and short-term memory networks are used to describe user interests,and a recommendation model based on user interests is designed.Through the analysis and calculation of user information,the LSTM is then used to describe the user’s preference time sequence to realize the accurate portrait of the user,complete the final recommendation,and avoid the phenomenon of solidification of the recommendation result.Secondly,the cold start problem in movie recommendation will cause inaccurate recommendation results and low user satisfaction.To solve this problem,this paper designs a recommendation model based on user relationships.The model uses the user information as an entry point,uses the user’s feature vector,analyzes and calculates the degree of interest influence between the target user and"friends",and completes the new user information based on their relevance,and predicts the results of the interaction.Complete the movie recommendation,thereby improving the accuracy of the recommendation result and enhancing the user experience.Finally,based on the recommendation model given by the above research,a movie recommendation application system based on user interests and relationships was designed and developed.The system includes a login and registration module,a personal homepage module,a system management module,and a movie recommendation module.It basically realizes the application function of giving different users personalized movie recommendation according to their interests,and has the characteristics of simple operation and stable operation.Through the above research,the movie recommendation model based on user interests and relationships has been further expanded and optimized,providing theoretical technical support for accurate recommendation and improving user experience;at the same time,the development of movie recommendation system based on this model also laid a good foundation for its practical application.
Keywords/Search Tags:User interest, User relations, Long short-term memory network(LSTM), Movie recommendation system
PDF Full Text Request
Related items