A brief look at Data Mining software tools
by
Siva Ganesh
Massey University, Palmerston North

Software tools used for data mining must be able to handle a comprehensive suite of methodologies. Commonly used data mining methodologies include market basket analysis, memory-based reasoning, cluster detection or segmentation, link analysis, decision trees and rule induction, artificial neural networks, genetic algorithms and bundling techniques.

Furthermore, these software tools should provide the expert analysts and statisticians with a graphical, user-friendly point-and-click interface via a desktop. However, they must be flexible so that user-defined approaches (in particular, to incorporate new and innovative ideas) can be easily integrated into the system. In general, besides the easy-to-use GUI, software tools must have comprehensive data input and output options including scalability, visualization capabilities to present results graphically to improve comprehension, multiple platform compatibility, service and consulting support etc..

This talk presents an overview of some of the popular data mining software tools such as SAS/Enterprise Miner, SPSS/Clementine, IBM/Intelligent Miner, CART (& MARS), S-plus 2000, STATISTICS/Data Miner and WEKA. A brief demonstration of SAS/Enterprise Miner and one or two other software packages will also be attempted.

Date received: May 8, 2002