Semantic Information Retrieval Based On Special Area And Relevant Technology

This paper is based on the concepts and technologies of semantic web, which is applied to information retrieval on special area. This paper makes a research on the information retrieval basing on the concepts of semantic web and its technology. Within the paper, relevant vocabularies and relationships in stock area was abstracted and defined accurately by means of ontology technology to build Stock Ontology. Also, stock information retrieval system has been developed with the following three core modules: stock ontology for the user query extension, common retrieval technology for core retrieval module in system, semantic rank algorithm for result ranking module. Meanwhile, self-adaptive hot pot is used to improve the semantic algorithm and the framework of semantic web information service is also presented in the paper.Semantic web is currently popular served as the main stream of next generation of Internet technology, which, when used in information retrieval based on special area, can improve the accuracy of retrieval result. On the other hand, ontology is also fit to resue knowledge and improve interoperability between machines automatically. So in this paper, related technologies of semantic web and the concept of description logic was introduced and then Stock Ontology (SO in brief) was built and coded by OWL.Based on construction of SO, Semantic Information Retrieval System based on Special Area (SIRSSA in brief) is presented and developed, which can provide semantic query extension function to help users to produce query either in standard systematic conceptual pattern or in triple node way. After the result set being obtained, semantic ranking algorithm-OntoRank was used to rank the result document set. The algorithm can analyze the concepts or relationships in certain document and calculate the weight of every item in result set which is very helpful to re-rank the retrieval result semantically and provide users the most relevant retrieval set.To extend the above algorithm, the concept of self-adaptive hot pot was presented in this paper. It can be useful to collect information about user click frequency and reflect the information or web pages users have paid more attention to. By use of this way, investors could conveniently receive the maximum relevancy of hot pot set.In order to provide stock information service to investment organization, the retrieval interface was coded by web service technologies. And then, ABC model was used to descript stock text information in semantic level. Also, different geography information can be integrated in the system to provide service of stock business by ontology reasoning.The paper also studied framework of application platform basing on stock area integrated with text and quantitative analyzing functions. Meanwhile, stock information ubiquitous service platform was also presented.