Design And Implementation Of Urban Community Recommendation System

With the development of society,massive data brings convenience to people's life,but also brings trouble.For example,the massive housing data makes users easily fall into the mire of information overload and difficult to obtain effective data;And location resources have become an important factor for users to consider when buying houses,but it is difficult for users to associate their demand for location resources with the data of housing sources.This paper is faced with the above problems encountered by users to design and implement.The system consists of six modules,which are data account,query statistics,location resources,housing recommendation,housing comparison and personal information.The data account module shows the overall housing price and location resources;The query and statistics module provides the function of house source query for users;Location resource module provides users with location resource related services from different perspectives;The housing recommendation module recommends the housing according to the user information;The housing comparison module is used for comparative analysis of different housing sources;The personal information module is used for personal information management.In the listing recommendation module,based on the research of recommendation algorithm,this paper proposes a hybrid recommendation model of content-based recommendation and user based collaborative filtering,which combines the content-based recommendation algorithm and user based collaborative filtering algorithm in a cascade way,and combines k-means algorithm to roughly classify the listing data and construct the matrix calculate the user's interest preference,use the gray correlation analysis to get the correlation degree between the houses.Finally,NDCG and Coverage are used to compare the experimental results of the model and the traditional recommendation algorithm.It is verified that the hybrid recommendation model proposed in this paper can achieve 47.8% in NDCG evaluation index and 43.3% in Coverage,which has a great improvement compared with the traditional recommendation algorithm.Finally,through the test,the system can help users quickly get the information they want to find,recommend house source and so on,and effectively solve the problems of users when buying houses,which has practical value.