Data mining is a new developing technology for enterprise data and information integration. It can reduce the operation cost, increase profit, and strengthen market competition of the enterprise. Generally, there are two ways to establish a data mining application tailor to an enterprise: using business intelligence solutions and products available on the market, or developing data mining algorithms all by oneself. However, both of them are impractical in cost and time. The former one costs a lot, while the latter requires developers to be familiar with both enterprise business and data mining technology.
Software reuse is a solution to avoid repeated work in the software development. It is regarded as a viable approach to solve the software crisis and promote efficiency and quality of software production. As a kennel technique to support software reuse, software component technique gets increasingly wide attention. To fully make use of reusable component, and support mass component’s production, classification, search, assembly and maintenance, component library is very important. Applying software component technique to data mining, wrapping individual business modules of data mining in the form of components, and using component technique to achieve the organization, management and retrieval of the components, could greatly increase the reusage efficacy and quality, and decrease the cost and period of data mining application development. The demand of variability of data mining tasks could be met as well. And the
application of data mining technology can be broaden[1].
Through componentization of data mining algorithm, design of data mining component library and development of component adaptive and assembly technique, this paper presents a new component library system of data mining and implements varied core algorithms of data mining in the form of components, which significantly improves the efficiency and quality of developing data mining software. By assembling components to analysis module and seamlessly integrate them into industrial information system, the organic integration of data mining software and industrial information system forms industrial information decision analysis system to meet various application demands. Compare to the existing data mining software, it can provide data mining service with more humanity characteristic and higher success rate, which is more suitable for real-time decision analysis application software.
The rest of this paper is organized as follows. In the next section, we introduce the related work. The discussion of the design of component model is given in Section 2. Section 3 gives the classification and retrieval strategies of the component library. Finally, Section 4 lists some conclusions and discusses some areas of future research.