Research And Application Of Resource Retrieval System Based On Query Expansion And Clustering Technology

With the development of Internet and enrichment of Web resources, obtaining information by Web-based full-text retrieval system becomes an important part of people's daily life. And users pay more attention to searching information more exactly and more efficiently.In order to enhance information retrieval accuracy of resource on Web and analyze the semantic structure of resources, the thesis analyzes the characters of Chinese word-segmentation, which help to combine unigrams, bigrams with word-segmentation based on word-lists. To enhance efficiency of inquriy demand, the thesis introduces a method based on query expansion and adopts different search strategy, which effective and necessary to enhance precision and recall of Web information retrieval.If the query conditions too broad and vague, the results return to users would be tremendous and it would take much time to look over all the results. Therefor, we cluster the search results furthermore. We describe the cluster by using STC. Using classified description in search context, we reduce the feature dimension by SVD technology and cluster the results. In these ways, the real time cluster of resource retrieval can be implemented, and organization of search results can be improved.In the final part of the thesis, we introduce the design of the project, Shanghai Education Resources Information Retrieval System. In this project, some relevant technology and framework are achieved. Then, we experiment and analysis the different search strategies. The experiments prove that the searching algorithm improve pertinency and focus level of...