Research Of Semantic Representation Mechanism Of Unstructured Information Based On Ontology And XML

With increasing development of computer network and software technology, especially with that of Internet/Intranet, large amounts of information resources are provided to their consumers in the form of various unstructured instead of structured formats. The applied range and quantity of unstructured information are rapidly expanding. Therefore, it is a valuable research subject to construct a consistent unstructured information representation, thus lay the foundations for establishing a platform to process and search the information efficiently.This thesis analyses the principle and advantage of using XML technology to describe unstructured information uniformly, aiming at its semantic limitation the thesis then imports the notion of Ontology and introduces a semantic representation mechanism of unstructured information based on Ontology and XML.XML possesses data pattern representing method, it has rich features such as separation of content and relationship, syntactic and semantic, content and representation etc. So it is very suitable to describe unstructured information. In spite of these positive features and prospects of XML, it must be clearly stated that XML is solely a description language to specify the structure of documents and thus their syntactic dimension. The XML tags and document structure can represent some semantic properties but it is not clear how these can be deployed outside of special purpose applications.Ontology is a formal, explicit specification of a shared conceptualization. Its target is to capture related domain knowledge, provide the shared understand of the domain, ascertain the shared vocabularies or terms of the domain, and make explicit definition about these terms and their relationships from different form levels. Our semantic representation mechanism imports semantic model from Ontology into XML documents by connecting Ontology with XML schema-level specifications, i.e. DTD and XML Schema, consequently, XML tags and structures can express explicit domain knowledge. Such a mechanism lifts information from the syntactic or representational level to the more abstract level of concepts and relationships, and avoids semantic heterogeneity efficiently. Furthermore, it provides existing applications a semantic-level solution about management of unstructured information, which facilitates these applications and enhances their practical value to a certain extent.Finally, the thesis realizes partial content of the semantic representation mechanism, and applies it to the OBSA model of a project named "Web storage system based on XML", which staked by the Hubei Provincial Department of Education organically. In OBSA, the mechanism promotes the information sharing and reusing, enables the semantic interoperation such as information search, information exchange, etc.