Research On Applying OWL Ontology To Manufacture Knowledge Management System

ERP system can improve the process of the enterprise operation and the competition power, but huge ERP system's structure is not agility and hard to be reused. By decoupling core product knowledge of ERP system, such as MRP (Material Requirements Planning), in models, not in code, it can be repurposed by the many different kinds of ERP systems that might require it. In this way, the efficiency of ERP system can be reformed. Furthermore, in the process of stock, manufacture and sale, there are complicated relationships between all kinds of information, so the relational database is not enough to describe the hiberarchy and interdependence interrelation, which can only be understand by human. In addition, the process of operating data can't be handled automatically and needs the help of handiwork. Thus we need a new flexible data-storing structure through which we can express all the semantic relation. Ontology can satisfy all requirements, as it is used to capture knowledge about some domain of interest in a computer-readable form to enable its exploitation, sharing and reuse.We use OWL DL as the language of building OWL ontology. Protégéand its OWL and PAL plugins are used for ontology development of concepts and relationships of mechanism domain. Firstly, we analyze the results of investigating the manufacture practice, and find some difficulties of expressing the data relations. For solving these problems, we apply OWL ontology to ERP system. We utility an OWL ontology approach to support mechanical parts knowledge storage and facilitate information retrieval. This paper mainly focuses on demonstrating the way in which an OWL ontology is built through using a renovated methodology, and the ability to infer new facts with a description logic classifier, RACER. And then we develop a tool to generate the PAL programs automatically, so the users need not to learn the PAL syntax. This way eases the stock of raw and processed materials and the sale of the products, and optimizes the enterprise repertory.We also show how this ontology can be reused and shared to perform ontology-based search with agents in Semantic Web. Finally we prospect our future work.