The Research Of Knowledge Integration Based On Ontology For Manufacurng Enterprise

With the rapid development of the world economy, enterprise enter the era of the knowledge economy. How to organize, obtain and adopt the knowledge, it will be the key that the manufacturing enterprise keep their competitive ability in the world. This thesis analyzed the problem from the application of knowledge interaction in the computer information technique and the artificial intelligence technique field, and combined the project from Gansu Science and Technology --"The Nonferrous Metals Industry Integrated Information System Architecture" (Number 2GS052-A52-003-11). Analyzing the deficiencies of enterprise knowledge integration, our group found that the system of knowledge is not shared and consistency the manufacturing enterprise, machines cannot understand meanings of knowledge correctly because of imprecise semantic description for knowledge. It is impossible to realize automatic and intelligent process of knowledge and so on. Applied the theories and methods of ontology, this thesis researched some key technology of enterprise knowledge integration and knowledge integration system.The integrated knowledge technology in manufacturing enterprise has several process methods. In this thesis, firstly, we described the structure of the knowledge ontology, established the theory and methods of Ontology, and provided the basis of knowledge integration for enterprise. On this basis, The Nonferrous Metals Industry Integrated Information System Architecture was realized in specific knowledge for enterprises, and the ontology model was established in the system. The systems choose the OWL language with the ability of ontology semantics as the express language to describe the whole ontology of the product knowledge in the manufacturing enterprise as instance.Aiming the problem of the integration of heterogeneous systems knowledge in ontology, the thesis researched one of the key integration technology -Ontology mapping, include its concept, principle, architecture and process. The method of the concept similarity calculation was improved. According to the examples, concept name and properties calculated the concept similarity, then combined all these by adding weighted proceed a new kind of similarity calculation method, finally tested the method.In the end, the main research result and innovations were summarized in the whole text, and the remaining problem and research works were analyzed in future.