Research Of Semantic Vertical Search Engine Based On Ontology

As an important means of accessing massive information on internet, search engine become an important field of information retrieval. With the rapid development of information technology, the quality of the internet information resources becomes larger and larger. Therefore, how to optimize the search engine in this area is concerned. As a professional search engine for a particular industry, vertical search engine is a subdivision and extension of search engine. The nature of the general search engines decides they can't satisfy the precise information needs service of special populations in the special areas. This paper studies ontology and vertical search engine, gives a model of semantic vertical search engine based on ontology, creates a method of semantic query expansion and optimizes the Lucene results sorting algorithm.This paper studies the basic concept and design theory of vertical search engine, ontology knowledge and its application in information retrieval. This paper analyzes the current status and deficiency of the domestic and international search engine and the significance and necessity of combining the ontology with the vertical search engine. This paper makes a contrast on the methods of building ontology and builds the digital products domain ontology, and use the Protege3.4.5 to complete it. And then, this ontology is used for division the user's query into ontology concepts, ontology instances, ontology properties and ordinary query. This paper focuses on the query expansion of ontology instances and ordinary query. For ontology instances, expand them from the two aspects of inheritance and path associate. For ordinary query, expand them based on word co-occurrence. Give a semantic query algorithm based on ontology for vertical search engine. In order to overcome the defects of Lucene sorting algorithm which only consider the related degree of the key words and web pages and does not reflect the importance of the web pages, introduce the PageRank algorithm based on link analysis to optimize it.Finally, develop a semantic vertical search engine based on ontology (DPOVSE), did an experiment to compare with the original query method and sort algorithm. The experiment result shows that the content of this paper's study can improve precision of search engines.