A web-aware interoperable data mining system

摘要

The development of web-aware data mining systems has received a great deal of attention in recent years. It plays a key enabling role for competitive businesses in the E-commerce era. One of the challenges in developing web-aware data mining systems is to integrate and coordinate existing data mining applications in a seamless manner so that cost-effective systems can be developed without the need of costly proprietary products. In this paper we present an approach for developing an interoperable web-aware data mining system to achieve this purpose. This approach applies Remote Method Invocation and high level code wrapper of Java distributed object computing to address the issues of interoperability in heterogeneous environments, which includes programming language, platform, and visual object model. The effectiveness of the proposed system is demonstrated through the integration and enhancement of the two well-known standalone data mining tools, SOM_PAK and Nenet, and runs with the iris data and air pollution data.