Design And Implementation Of News-Oriented User Behavior Capture System

With the explosive growth of Web information, how to quickly and accurately get the information we needed from the vast resources has become a problem. Although search engine technology resoulve the problem for some people, but because of its universal nature, still can not meet the different backgrounds, different purposes and different times of the users' personalized needs. User information collection is to address the issue. Capture user behavior research goals is to fully utilize available to the user's personalized information, guided by user interest or user interaction to provide users with the information users needed, also make use of information, to serve the user's individual requirements more better. The collection and analysis of user behavior, whether in the availability of Web site design,or personalized for the user modele, are both great significance.In this paper, we implemented a subsystem, which is part of prototype system the news-oriented automatically recommend.That is the system of news-oriented user behavior capture. I complete the requirements analysis and design with other team members,and then I do the work of the system implementation and testing by myself.In the field of news,we implement the prototype system based on j2ee, according to the register information when the user using the system, customized user profile. And then capture the user's action of Web browsing, retrieval of information by implicit way,.This information will be processed mapped to the user's interest value in different news categories.At last using the data handled continuously update and maintain the user's personality model.On the other hand, to resoulve the problems of search engine technology, in order to capture the users'search behavior more accurately. The paper analyzes ontology description language, ontology algorithms, ontology modelling, reasoning rules and methods based on the ontology in details.In addition, catering-oriented model based on ontology has been eastablished with relevant ontology techniques, such as ontology modelling based on Protege, semantic inquiry based on Jena API and reasoners based on Racer. Finally, to evaluate and verify the correctness of the results, three typical examples of concept retrieval systems are discussed. And with this, through the ontology and search technologies to achieve an organic integretion, then enhance the users'retrieval precision rate and recall rate.