Research And Realization Of Personalized Services Of Intranet

With the rapid development of Internet, network has become a kind of necessary in people's daily life. However, along with continuous extension of the network and increase of information quantity, the users discover that it is more difficult in getting information as correctly and quickly as they need. It because traditional information service system is passive and basing on manual searching, which is unable to meet the growing diversity and individuality of the information under the circumstances of "information overload". Then the personalized information service mode is brought out which can automatically organize and adjust the information according to users'interest. Personalized information services mode is an inevitable requirement with the rapid growth of Internet and has become an urgent and important question for study.The thesis provides an in-depth study and research on the background, feature, performance and key techniques of the personalized information services mode, and the analysis of the status of Intranet. an innovation will be achieved on the basis of combining the study of a personalized information service with the characteristics of Intranet. A personalized information service system is designed and realized. The followings are the key contents of the thesis.1,Analyze the status and the limitation of the Intranet, provide specific goal of personalized information services. The current information channels provided by Intranet apply traditional manual-searching modes. To meet the needs of employees for the personalized service and improve the information searching efficiency, personalized information services should be introduced to Intranet.2,Study the key techniques of personalized information service, including web data mining technique, users modeling and personalized recommend technique. According to the different objects of data mining, users'participation level in modeling procedure and the different realizing methods, we study the techniques of web content mining, web structure mining and web application mining, and practice of the manual modeling, sample modeling and automatic modeling and the recommend technology based on rules, contents and Interoperability filter. 3,Conduct an in-depth research on the web log data preprocessing and provide correlative data pre-processing algorithm. The experiments prove accuracy and validity of the data process in the thesis and provide the data source for extracting users interest.4,Present an Intranet users interest model is provided which integrates static point-of-interest with dynamic point-of-interest. The concept of role is introduced in Intranet and the core algorithms of personalized information service system are provided. Staffs holding the same posts are regarded as the same type of roles according to the features of enterprise management.5,Put forward an Intranet users interest model, a roles interest recognition algorithm and page recommend algorithm basing on roles interest, which are successful used in experiments.6,Design and realize a personalized information service of Intranet system .The practical testing on Intranet proves that the system can effectively capture the users'point-of-interest and can take the initiative to recommend information to users based on their point of interest. It also shows that the system can improve the browse efficiency and increase resource utilization ratio and ultimately realize the personalized services.