In order to better achieve knowledge sharing based on distributed ontologies, an approach based on ontology context immigration （OCI）is proposed. Compared with traditional approaches such as ontology integration and...In order to better achieve knowledge sharing based on distributed ontologies, an approach based on ontology context immigration （OCI）is proposed. Compared with traditional approaches such as ontology integration and mapping, the proposed approach can reduce the implementation complexity. This approach can be mainly divided into three phases： ontology context determination for a given term, ontology semantic similarity computation between ontology terms, and ontology context immigration. As for a local semantic term based on distributed ontologies, an appropriate ontology context of the term is determined and extracted from a local ontology most associated with the term by using semantic similarity computation. Then, the ontology context is dynamically immigrated to the source ontology for enriching semantic information related to the term. A system called distributed knowledge sharing system（DKSS） is developed to illustrate this approach. The system adopts multi-agent technology for better communication and coordination between different ontology information sources. The experimental results show that it is efficient for distributed ontology knowledge sharing. The proposed approach does not require the support of a global ontology or the maintenance of complex ontology mapping relations, and thus it has better maintainability and scalability.展开更多
基金The National Natural Science Foundation of China(No60703036)
文摘In order to better achieve knowledge sharing based on distributed ontologies, an approach based on ontology context immigration （OCI）is proposed. Compared with traditional approaches such as ontology integration and mapping, the proposed approach can reduce the implementation complexity. This approach can be mainly divided into three phases： ontology context determination for a given term, ontology semantic similarity computation between ontology terms, and ontology context immigration. As for a local semantic term based on distributed ontologies, an appropriate ontology context of the term is determined and extracted from a local ontology most associated with the term by using semantic similarity computation. Then, the ontology context is dynamically immigrated to the source ontology for enriching semantic information related to the term. A system called distributed knowledge sharing system（DKSS） is developed to illustrate this approach. The system adopts multi-agent technology for better communication and coordination between different ontology information sources. The experimental results show that it is efficient for distributed ontology knowledge sharing. The proposed approach does not require the support of a global ontology or the maintenance of complex ontology mapping relations, and thus it has better maintainability and scalability.
