A CROSS-COMPARISON MATRIX OF DATA MINING SOFTWARE SUITES FOR USE IN UNDERGRADUATE DATA MINING EDUCATION

Abstract

During the ten years from 1993 to 2003, datamining software was commonly comprised of a loosecollection of algorithms and tools, with limited common userinterfaces and inconsistent outputs. These software productswere created primarily in the computer labs of graduateschool computer science programs, gradually merging intobusiness school curricula. This natural progression in datamining software has led to the development of data miningapplication suites, bringing diverse tools for categorizationand prediction into a single location with intuitive, graphicaluser interfaces and common outputs. The use of a singlesoftware product in teaching data mining has now becomemore practical, and has enabled the teaching of data miningconcepts and techniques to non-traditional studentaudiences. This paper examines a feature comparison of themost common data mining suites available for use in theclassroom, including costs, management demands, andusability. Results and recommendations are reported.

Index Terms - Data Mining, Education, Software.