| Ontology is specification of shared conceptualization and is building block of the semantic Web. Ontology building requires a detailed domain analysis that in turn requires financial resources, intensive domain knowledge and time. Most of industrial data is present in relational databases and a relational database schema represents a domain model. An ontology built from this schema can represent concepts and relationships that are present in domain of discourse. However, databases are not static and their schema evolves over time. Once a database schema is changed, these changes in schema should also be incorporated in ontology, generated from this database. The possible solution of regenerating a new ontology from changed database schema is not feasible because this will result in loss of manual changes of ontology. In this paper we present an approach that can be used to generate ontology from RDBs and to synchronize the generated ontology with changes occurred in the same database.This ontology has been used to guide the design of the application and to supply the system with semantic possibilities. The use of ontologies helped to model the domain in a reusable, shareable and more efficient way. The restrictions included in the ontology facilitated the system capability for predetermining allowed values for most attributes. The use of ontologies as a modeling instrument implies other important advantages. It allows the system to check, automatically in most cases, for the consistence of the information introduced by the user, so preventing it from accepting information inconsistencies.There are two issues of domain ontology building and evaluating:1) Ontology construction is apart from ontology evaluation;2) current methods more or less overlooked the role of ontology end-user. This paper proposes a Web-based ontology development framework, with domain experts, ontology engineers and end users to participate in it, which emphasizes on the continuity between ontology building and evaluating process and the importance of user-centric. Its structure, work principle and key factors about implement were particularly discussed. Compared with the past ontology development method, the framework makes full use of information diversity, functional diversity and breadth of audience of the Web, and emphasizes the user-centered concept of ontology development. The proposed algorithm solved the problem of low efficiency and inferior precision that calculating the similarity only depended on the attribute value and did not consider semantic and attribute type influence in the traditional recommendation algorithm. Moreover, the semantics of the used languages for expressing the domain knowledge is formally defined in a non-ambiguous and well-understood manner, e.g. in form of some variant of First Order Logic. The clear separation of domain knowledge and inference knowledge allows us to build easy-to-maintain knowledge bases.The creative work in this thesis is as follows:1. Build a domain ontology of goods;2. And the domain ontology applied to information retrieval, to achieve the purpose of the intelligent guide users shopping;3. On this basis, intelligent search system based on the domain ontology of goods has been implemented.This thesis is the project supported by the Guangdong Provincial Natural Science Foundation (grant no.8151009001000041)——the research of the "Software agent strategy generation mechanism research"... |
PDF Full Text Request