Author(s)

J.G. le Roux & H.L. Viktor

Abstract

A cooperative data mining team consists of two or more diverse data mining tools that cooperate to learn. The data mining tools learn from one another by accessing the other team member's knowledge, generating new data and thus adapting its own perspective of the problem at hand. By combining the complementary learning approaches followed by diverse tools, new insights into the problem domain are obtained. This paper describes the ViziMine cooperative data mining tool that acts as a personal assistant, assisting domain experts to explore and increase their expertise regarding a problem domain. The tool provides a uniform interface that incorporates a number of data mining tools that express their knowledge as rules. Results indicate that the approach successfully facili- tates the cooperative learning process and eases the transfer of k

Keywords