Conceptual Analysis Of The Query And Its Application In Retrieval

With the development of computer technology and the rise of theInternet, information resources in the network had been rapidly growing. Itbrings massive amounts of information, while at the same time, it also bringsthe problem of information explosion. Facing the complicated networkresources, the most concerned problem is that how to get candidatedocuments from the vast amounts of information retrieval system which canbest meet users needs.Information retrieval system, at present, can only provide limited help tousers. Besides its low accuracy of retrieval, some candidate information is nothelpful, but brought little troubles. The crux of the problem is that mostexisting retrieval systems using discrete model, like Boolean model. Theneeds of users and documents are represented as discrete string, the words areunrelated to each other, as a result, it loses their conceptual integrity, andbrings a new noise. A feasible solution to this problem is to introduce meansof natural language understanding into retrieval, using deep-level semanticanalysis to improve the retrieval accuracy. It is that applying semanticsanalysis to indexing needs and documents, and the basic indexing unit iscomplete concept, replacing of string. Thus, it can build the relationshipbetween concepts of needs and documents.This paper is about conceptual analysis of user needs in Chinese, whichis the essential part of Chinese concept retrieval system. Needs analysis is thefirst step in the retrieval process, its purpose is to restore users’ search intent to guide the further search. Thus needs analysis is the primary task of theretrieval system, more important to accurately portray retrieve concept inthe user’s mind. This is based on users’ need expression, which may be fuzzy.This paper tries to use semantic conceptual graph model to processChinese query on a conceptual level, and as a result, the query will betransformed into a semantic concept map. It will link words entered by theuser, into the complete connotation graph based on the semantic relations.Thus it will not lose its semantic concept information during the wholesemantic retrieval process. As a result, retrieval system will return candidatedocuments which may more meet users’ needs.This paper proposes a new method, which is to analyze users’ needsfrom the view of Chinese conceptual connotation, with the method ofsemantic analysis, to restore the users’ retrieval intention, and then to guidethe next search. It will get the focus information, which should replacewh-words, and finally build the conceptual graph which contains theconnotation of semantic information of needs. This new method may provideeffective technical support for the future development of a new generation ofChinese search engine, and it is an important part of language engineering.