Design And Implementation Of News Recommendation System Based On User Feedback Information

The choices of network news become more and more with the coming of theinformation age. In order to recommend the network news which the users interestedin to the them and develop one personalized recommendation system for news whichcould help users obtain their interested information so that it can help users greatlyreduce their time on searching for news, the personalized recommendation system fornews become a significant content in the research of network news retrieval. Based onthis, the paper mainly carry out the work from the following several aspects.(1)In the initial stages of designing paper, this paper expounds thoroughly theresearching background and significance of the paper by studying a large number ofrelevant documents on the recommendation algorithm and preliminarilyunderstanding about the personalized recommendation system for news. What’s more,the functional requirements of the system and several major functions: InformationCollection, Data Preprocessing, Topic discovery and News recommendationdetermine have been designed. And on this basis, the system is operating environmentconstraints limit on nonfunctional requirements and performed requirements.(2)Recommendation algorithm is the most core and key part of therecommendation system. In this paper, the algorithm of personalized recommendationsystem for news based on the users’ feedback information was designed and appliedinto the system by studying the existing recommendation algorithms. Algorithmconsists of three basic steps: Deal with the document with the TF-IDF method;Calculate the Topic similarity; Calculate the recommend index.(3)Based on the system requirements analysis, every aspect has been designed indetail. We made the following design. By comparing of the advantages of C/S modeand B/S mode for this system, the system design to adopt the B/S mode ofdevelopment. At the level of the system design, in order to realize the characteristicsof high cohesion and low coupling, the system architecture of the system can bedivided into three layers: Data access layer (DAL), Business logic layer (BLL), andPresentation layer (UI). In the personalized recommendation system for news based onthe users’ feedback information, three modules, including data acquisition and topicdetection and news recommendation were designed; In addition, the system selectedthe database version of the Microsoft SQL Server2005as the database developmentenvironment, the technology of C#as development language.The personalized recommendation system for news based on the users’ feedbackinformation has realized the function of information acquisition of network news, thenews data processing, hot topic detection and hot topic founding. And the data will bepresented to the user according to user’s feedback information, and made thepersonalized news recommendation possible. By the system testing, the personalized recommendation system for news based on the users’ feedback information can be putinto practical application.