Based On The Concept Tree Pruning Of Lca Query Expansion Technology Research

With further infrastructure and popularization of internet service, information explosion has been emerging as a major feature to the information society. The latest research revealed that about84%of the computers users will conduct to using different search engines in their daily life, more than half of them are on a daily basis. The information has never been produced and released as fast as they have now and the internet users demand a developing information retrieval capacity, as have formed a great challenge to nowadays information searching tools. Search and find out useful information from a mass stock in a very short time is becoming a tough tax and many times the user will have to receive a lot of information that may not appear any useful for the users. Problems mainly come from the query which users have provided is not accurate enough, the mismatches of the queries and the expression of the documents, or during query optimization. To deal with these problems, query expansion technique appeared on the scene. By expanding the initial query, query expansion technique improves the performance of information retrieval, which has important theoretical significance and practical value to research.This article has been composed by following parts:First、This paper analyzes introduce the ultimate objective of the thesis, the basic conception of the query expansion and the main searching process and so on. And we also give a study summarization on information retrieval technology both at home and aboard. Currently, four query expansion techniques have extensively applied, namely, Query expansion based on global document analysis, Query expansion based on local analysis[1-4], Query expansion based on query log analysis and Query expansion based on association rules. A comparison of the four searching methods has also been given in terms of their basic guideline, the difficulty degree of the algorithms and retrieval performance.Second、Based on the existing query expansion technologies, we establish a simple effective and feasible technology-the Local Context Analysis, which as the main expansion method in this paper. Many researchers in the technology have made gratifying achievements. But the method strongly related to the dependence of the similarity between the initial retrieval results and the initial queries. If the similarity between the initial inspection results and the initial queries is not high, a large number of unrelated words will added into the initial queries, then, the precision ratio will severely reducedThird、This article discusses the query expansion based on semantic resources, which expanded word from the descriptor list (Here we choose the canonical terms from their Synonyms, Hypernyms, Hyponyms and Holonyms etc.). With the lexical semantic such as WordNet, HowNet, we construct the semantic tree, which shows a more intuitive semantic query expansion result sets. This technology is superior with accuracy comparable to the Local Context Analysis.Fourth、This paper proposes a novel hybrid query expansion method which synthesizes the merits of semantic query expansion and local context analysis (LCA). Firstly, we retrieval the documents by LCA method, then use these terms to trim the semantic tree, and calculate the weight of expansion term based on this improved algorithm. In this paper, we compare the effectiveness of these approaches and show that, although local context analysis has some advantages, the LCA prune semantic tree yields better performance than the techniques on the simple query expansion.