CAJ | 학술논문

为了更好地实现基于分布式本体的知识共享, 提出了一个基于本体环境迁移(OCI)的方法. 该方法同传统的本体集成和映射方法相比, 能减少实现复杂度. 该方法可以分成3个阶段: 给定术语的语义环境确定, 本体术语之间的语义相似度计算以及本体环境迁移. 针对分布式本体的一个本地术语, 其本体环境可以通过使用语义相似度计算从与该术语最相关的本地本体中确定和抽取. 然后, 该环境将被动态地迁移到源本体以获取更丰富的语义信息. 采用多智能体技术, 开发了一个分布式知识共享系统(DKSS)以演示该方法的使用. 实验结果显示, 该方法对分布式本体知识共享是有效的, 不需要维护全局本体或复杂的本体映射, 因此具有更好的可维护性和可伸缩性.
위료경호지실현기우분포식본체적지식공향, 제출료일개기우본체배경천이(OCI)적방법. 해방법동전통적본체집성화영사방법상비, 능감소실현복잡도. 해방법가이분성3개계단: 급정술어적어의배경학정, 본체술어지간적어의상사도계산이급본체배경천이. 침대분포식본체적일개본지술어, 기본체배경가이통과사용어의상사도계산종여해술어최상관적본지본체중학정화추취. 연후, 해배경장피동태지천이도원본체이획취경봉부적어의신식. 채용다지능체기술, 개발료일개분포식지식공향계통(DKSS)이연시해방법적사용. 실험결과현시, 해방법대분포식본체지식공향시유효적, 불수요유호전국본체혹복잡적본체영사, 인차구유경호적가유호성화가신축성.
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.