The Future of Data Mining

다국어 초록 (Multilingual Abstract)

Data mining has come of age in recent years. The technology of data mining - covering machine learning, statistical approaches, and visualization methods - now represents an indispensable toolbox for decision making in business, government, engineering, and science.
A competent software system should possess greater functionality than merely fetch data or broadcast simple information. In particular, an intelligent software agent should be able to perform routine tasks autonomously, glean new knowledge from disparate databases, and improve its own performance through experience. Such capabilities can be implemented through the techniques of knowledge discovery.
Where are the opportunities for research and implementation of data mining tools over the next decade? The most promising fields for research appear to lie in greater intelligence as well as the embodiment of data mining techniques within autonomous agents. Moreover, such agents will take perceptible form through multimedia projections and mobile agents which navigate autonomously through virtual worlds on the Internet. This paper examines a number of critical issues behind the development of such agents, formulates a general architecture, and demonstrates the deployment of a learning agent for industrial planning.