Research On Key Technology Of Meta Search Engine Based On Semantic

As the continuous development and improvement of Internet technology, the information in the internet shows explosive growth. In order to obtain valuable information from a large number of resources, the search engine came to being. But facing the characteristics of Internet document, such as dynamic nature, no structured and semi-structured, the search engine exists limitations of little coverage and long update cycle of the database. In order to give full play to the search efficiency of the independent search engine, Meta search engine which by means of independent search engine as a child retrieval tool becomes the focus of one of the areas of researcher's concern.Based on research of Meta search engine and Ontology technology, this paper found that can use the characteristic of Ontology knowledge expression to improve the intelligent understanding of user queries. Moreover Ontology is a conceptual model and a clear formal specification. Therefore, it is a great practical significance and application value that apply ontology to Meta search engine to achieve functions of the semantics of user's query processing and understanding. The focus works of this paper are as follows:(1) The system structural model of Meta search engineA Semantic-based Meta search engine system model is designed and implemented. The model framework can obtain subject information from the network which has non-structural and semi-structured documents. And utilize the subject information to carry out ontology annotation. Under the premise of suitability, combination of the semantic features of Ontology and the advantages of Meta search engine, the model can improve effectively user's satisfaction and the relevance of search results.(2) Pre-processing algorithm of query expansionAccording to the problem of semantic relevance for the query keyword is not high, this paper proposed a pre-processing algorithm of Meta search engine based on semantic. Between utilizing Ontology's hierarchical structure, calculating the semantic concept level tree's relevance, and introducing the depth restriction function in the concept similarity's computation, then through with the Ontology knowledge base's mapping, reaching its ontology-based semantic extension. Experiment show that the algorithm greatly increased the numbers of hit target page, and improved the correlation of Meta search engine query results effectively.(3) Results processing algorithmBased on previous research, by integrating the result fusion technology into the Ontology based four levels of results processing. Therefore this paper proposed a results processing algorithm. The algorithm uses a series of measures that includes duplicated web pages, introduced lemma-match grade, calculate the correlation degree and sorting method and so on, to realize to the target of detailed processing of the returned results. Through with the experiment and analysis show that the algorithm can improve the user queries related degree effectively, and satisfy user's real inquiry intention.(4) Analysis and VerificationBased on the semantic Meta search engine model, the pre-processing algorithm in pretreatment module and the results processing algorithm in results processing module are verified and analyzed separately. The experiment indicated that the two algorithm based on the model structure has the feasibility and effectiveness in the time efficiency and user satisfaction degree.