Research Of Query Expansion Based On Domain Ontology

Query Expansion technology, which can improve the accuracy of Information Retrieval, has been widely applied to the Search Engines, and becomes a research hotspot. Ontology is a clear formal specification for the shared conceptual model, which includes four meanings: the concept model, definitude, formalization, sharing. Ontology can be used to describe the concepts in one or some fields and the relationships between the concepts, which makes the information during sharing have acknowledged, specific and only definition. So Ontology can help people efficiently retrieve useful information in a standardized information system.In order to improve the Query Expansion, this dissertation advance a new Query Expansion method based on the Domain Ontology, combined with some advantages of the normatively and the clear hierarchy in the Domain Ontology. This method overcomes shortcomings of the traditional Query Expansion technology, such as bad semantic and bad topicality.The main research works of the dissertation are summarized as following:1. Advance a novel method of extracting the Field Association (FA) terms based on the Domain Ontology. The Domain Ontology is built based on the Google page directory. The pages from which the Filed Association terms need to be extracted were segmented, and the results of these terms formed a new set of these terms. The Field Association terms which satisfying the given threshold value are extracted from the set based the Domain Ontology. These Field Association terms can describe one field or several fields. These Field Association terms are filled to the existing Domain Ontology, and form new Domain Ontology. The experiments show that the Field Association terms extracted by my method are superior to the key words extracted by the traditional methods in express documents.2. Propose a method for forming Query Expansion Source based on Formal Concept Analysis. A concept lattice is built by using the big data set as the context. Then compute the LUB and GLB for the uses' query terms. And then, we make sure the Query Expansion Source in the concept lattice, based the users' query terms and their LUB and GLB.3. Propose a novel Query Expansion method based on the Domain Ontology. Match the terms of the Query Expansion Source and the concepts of the Domian Ontology. Make sure the matching concepts and the matching attributes in the Domain Ontology. Based on the existing Domain Ontology, build a Sub-Domain Ontology centering on the matching concepts and matching attributes. And then achieve the generalization and refinement for the users' query terms. Ultimately achieve the users' query expansion. Comparing the query expansion results of my method and the other Search Engines, the experiment indicates my method is feasible.