Personalized Web Information Retrieval System, Design And Implementation

With the tremendous growth of Internet, web information forms a global distributed complex information space. In order to help the user to retrieval the useful web information, the, web information retrieval system comes into being. The web information retrieval becomes a new research and development field, and the new developing trend is the intelligentized retrieval and vertical retrieval.The thesis introduces the theory of the web information retrieval and the mathematics model in IR, and based on analysis and discussion of the Web information extraction, word segmentation and user-machine interface technique designs the personal web information retrieval system with sound function, reasonable structure and practical feasibility. The system gets the Web information through meta-search engine, introduces the multi-stage retrieval approach on Riemann space model, ranks the documents according to their geodesic distances to the query, at the same time introduces the user modeling creating the users database to realize retrieval personalization. The system is applied in the practical project" The Research of Chinese Human Brain Project & Neuroinformatics ": designing and realizing the personal neuroinformatics web information retrieval system using the Riemann space model and user modeling technology. The program is written in Java in Windows 2000 server and the database is Oracle 8i to save the professional dictionary, feature vector and user log. The system is on trial and presents good performance.The study of the thesis settled the retrieval problem of information about neuroinformatics, and is universally valuable for the design and realization of web information retrieval. The thesis frames and implements personal specialty web retrieval system. The system is schema for the generalization of vertical retrieval and has wide study meaning and practice value.