A European Commission-funded industry consortium which has grown to include leading US players as well, yesterday announced it had completed its project to develop a methodology for data mining. Crisp-DM, the Cross Industry Standard Process for Data Mining, is the product of a project instituted under the European Commission’s fourth Esprit cooperative IT research framework, and originally proposed by NCR, Daimler-Chrysler (formerly Daimler-Benz), ISL (now part of SPSS) and OHRA, a major Dutch health insurance company.
Colin Shearer, SPSS’ VP of data mining business development, said the project has garnered the support of most the world’s major data mining software companies, and a great many of its largest users in the course of the 18-month project. There are now more than 190 organizations from around the globe in the Crisp special interest group (SIG) he said.
The completed Crisp method, or process model, will make it easier for data mining users to replicate their data mining projects, as well as providing a benchmark for correct practice which will help convince more companies that data mining is now a mature technology, said Shearer. He and other members of the Crisp project now hope to persuade their fellow members to turn the SIG into a subscription membership organization, with a view to publishing the Crisp model before the end of this year, and then perhaps to carry the work forward and have Crisp recognized by other standards bodies, such as the International Standards Organization.
