A methodology for evaluating and selecting data mining software

Abstract

As data mining evolves and matures more and more businesses are incorporating this technology into their business practices. However, currently data mining and decision support software is expensive and selection of the wrong tools can be costly in many ways. This paper provides direction and decision-making information to the practicing professional. A framework for evaluating data mining tools is presented and a methodology for applying this framework is described. Finally a case study to demonstrate the methods effectiveness is presented. This methodology represents the first-hand experience using many of the leading data mining tools against real business data at the Center for Data Insight (CDI) at Northern Arizona University (NAU). This is not a comprehensive review of commercial tools but instead provides a method and a point-of-reference for selecting the best software tool for a particular problem. Experience has shown that there is not one best data-mining tool for all purposes. This instrument is designed to accommodate differences in environments and problem domains. It is expected that this methodology will be used to publish tool comparisons and benchmarking results.