Applying Ontology In Data Mining Knowledge Management System

Data Mining technology is the key field of computer science. Because of the continuous expansion and update of the knowledge in this field, some problems also appeared, such as: 1. It can not realize automatic defining and classification for new domain knowledge, so it needs the manual manipulation of experts. Therefore, the man-made discrepancy of definition and classification appears. 2. There is no unified data mining knowledge management system for users to search knowledge. 3. When an ordinary user who doesn't have much domain knowledge submits a DM task, an optimal combination of DM method and algorithm could not be obtained possibly to solve this task.In order to solve the above problems and assist the work of data mining knowledge management, it is necessary to conduct intelligent management for DM domain knowledge. Therefore, this research introduces Ontology into it. The major works include:1. This research analyzes the existing problems in the using of DM, points out the disadvantages of the present DM knowledge management and summarizes four primary problems. Based on this, it introduces Ontology and analyzes its functions and advantages.2. It presents an exoteric and extensible architecture for DM knowledge management and gives a description of the workflow of each component and the system.3. It mainly gives an illustration about the Ontology model of DM knowledge management, including the Concept, such as DM task, method and algorithm; Instance, such as the common used method and algorithm (Apriori, ID3 and BIRCH) instances about association rules mining, classification mining and clustering mining; Relation. Then it constructs a DM Ontology model using Protégé. Besides, it investigates the realization of simple deductions of knowledge concepts using PAL language and the axioms supplied by OWL.4. It conducts an experiment on DM knowledge management system, focusing on the realization of DM knowledge query and the intelligent match of DM task.5. At last, by concluding the whole research, it lists several problems which are worthy of further research. The significance of this research is that it puts forward the application of Ontology in the construction of Data Mining knowledge management system and constructs the experimental system. The results indicate: DM knowledge representation method based on Ontology can realize the semantic representation in DM effectively. The query system is intelligent in some degree and can provide valid support to the intelligent search of DM knowledge. It also verifies the feasibility of the data mining knowledge management architecture.