Design And Implementation Of Film Recommendation System Based On Graph Database

With the rapid growth of the Internet, information overloading has become a serious issue. Finding useful information from the Internet is more difficult than before. Search engine may help to deal with some aspects of the information overloading problem, however it couldn’t solve the issue completely. For example we really need a recommendation system to help us find valuable information when someone could not describe his or her requirements clearly. In multimedia area, users also have to face the fact that there are so many movies that they cannot find the movie they liked most. So the research of personalized movie recommendation is of great significance to the solution of information overloading problem in movie field.In this context, we proposed a hybrid recommendation algorithm which combine the classical collaborative filtering based on user-ratings with the weighted genre in a movie. Firstly, we analyzed the majority algorithms which are widely used currently, including content-filtering, collaborative filtering based on user or item, and compare their pros and cons. Secondly, we proposed a improved recommendation algorithm which given a weighted genre on a movie, analyzed history ratings of the user and mined his or her potential user preference depend on characteristics of the film itself. Finally, we designed and implemented a recommendation system which use the weighted genre on collaborative filtering algorithm, the software system utilized the dataset of MovieLens, programmed with Spring framework and neo4j and shown its result of recommendation through the Web.The algorithm proposed in this paper may has certain reference value for the movie recommendation system. The thought of given a weighted value on genre information can be used on other context information or other recommendation system. The final implemented system could be used as demonstration system of some relevant recommendation algorithm.