Research And Application Of Recommendation Algorithm Based On Two Part Graph Analysis

With the rapid development and popularization of the network people's life is closely related to the network,the information overload on the house market network becomes a serious problem.With the information about the housing rental,second-hand housing sales on the housing sites and intermediary sites getting more and more,some information can not be updated making it difficult for users to efficiently find their personalized housing information.The most important issue to the web site is how to provide users with their interest information in real-time and accurately.Personalized recommendation system is an effective measure to solve this problem.In this paper,the author participated in the work of 365 Amoy network recommendation module.According to the needs of the project,this paper studies the personalized recommendation method of the housing from the perspective of community mining,and designs the corresponding housing recommendation system.The main research work and results are as follows:(1)We propose a recommendation algorithm based on the community mining of the two graph.We use the weighted two graph to represent the user-item rating matrix.First,we divide the user into the community,then divide the user who needs to be recommended to the most relevant community,and then use the similarity between users to recommend.The algorithm takes into account the relationship between the user and the community,as well as the similarity between the user and the project.(2)The Jaccard_with_SVD(Singular Value Decomposition)algorithm is proposed in this paper.We combine Jaccard with singular value decomposition algorithm to improve the accuracy and recall rate.First,we use Jaccard algorithm to find the recommended listings index ranking matrix,which added an incremental update.Then we use the SVD algorithm to fill the Jaccard algorithm of the Index ranking matrix,so as to get a complete list of recommendations,according to the actual requirements recommended.The experimental results show that the proposed algorithm has higher performance compared with other related algorithms.(3)According to the need of the recommendation system of the 365 housing network,this paper designs and implements the recommendation system based on the recommendation algorithm of community mining and the hybrid recommendation algorithm of Jaccard and singular value.In this paper,we made a recommendation system for housing demand analysis,put forward the overall design of the system framework,and the framework design of multi module structure,realizes the function to the user recommended recommended housing.