Research On Semantic Retrieval & Its Semantic Similarity Based On Ontology Technology

With the development of network technology and rapid increasing information on Internet, information retrieval system plays an important role at communication between users and resource on the network. The traditional information retrieval is only based grammar match, which lack of the presentation, handling and understanding of knowledge. The key problem is that information resource is lack of semantic description, so that it is hard for users to retrieve the information which they really want and impossible to associate information resource with semantic feature. The essential solution to this problem lies in the information retrieval from the traditional grammar-based level upgraded to knowledge-based semantic level.Semantic Web is an extension of the current one, in which information is given well-defined meaning, better enabling computers and people to work in cooperation. Ontology has the good hierarchical structure of concepts and the support of logical reasoning, and semantic information can be realized through the semantic relationship of concepts. Ontology technology can be well applied to information retrieval. Ontology-based information retrieval is different from the traditional keyword search. Semantic Intelligent Information Retrieval can be realized because Ontology knowledge base strengthens the intrinsic link of the concepts and the implied and unclear information can be deduced through logical reasoning. This paper analyzed the traditional information retrieval technology and got that the reason of the low quality of its retrieval fundamentally lies in the traditional information retrieval based on the matching syntax and lack of the semantics of information retrieval. And this paper put forwarded the Ontology technologies to be applied to information retrieval. Another way, Ontology technology applied in the field of telecommunications applications was analyzed in detail, including Ontology-based network management system integrated information model, Semantic Web technologies in the context-aware smart mobile Web services and ontology construction in telecommunications field.Then this paper focuses on the analysis of several key technologies of ontology-based semantic intelligent information retrieval, including ontology technology, the method of Semantic Intelligent Information Retrieval, domain ontology building process, and system process. Based on analysis of traditional information retrieval technology and ontology technologies, Ontology-based Semantic Intelligent Retrieval System was designed. After analysis of the current information retrieval system of on-line mobile phone product shop on the Internet website, the semantic intelligent retrieval system framework model based on ontology was proposed. Then mobile phone product ontology was constructed for the experimental system, and the semantic reasoning was analyzed in Semantic Intelligent Information Retrieval.After that, Mobile Phone Product Semantic Retrieval System (MPPSRS) was developed based on the technology theory and system design in previous sections. Mobile phone product was the intelligent retrieval object in this experimental system. Through the semantic reasoning based on ontology, we can fully explore the retrieval of information which users implied. This system offered a good semantic retrieval services which fundamentally solve the shortage of traditional information retrieval in which information resource was lack of semantic information, and this system provided users the more accurate and comprehensive retrieval result as users' inquiries and achieved Semantic Intelligent Information Retrieval.At last but important two sections in this paper, traditional concept semantic similarity computation models was analyzed, and based on domain ontology, a reformative semantic similarity algorithm was put forwards, which integrated semantic similarity based on distance and semantic similarity based on attribute. For distance-based semantic similarity, several important elements which are implicated in domain ontology were taken into account, such as semantic ancestor, semantic depth, semantic distance, semantic density, related adjustment factors and so on. Then an ontology base from an actual company, Ballasts and Lamps, was developed and a semantic similarity retrieval experimental system, Ballasts & Lamps Product Retrieval Recommendation System (BLPRRS) was developed. And the experimental result demonstrated this semantic similarity computation model could help to extend the query concepts sets and provide an effective product retrieval result.