PublicationsFrom: Gregory Piatetsky-Shapiro gpsDate: Tue, 22 Aug 2000 06:52:18 -0400 (EDT) Subject: CRISP-DM 1.0 proposal published CHICAGO--(BUSINESS WIRE)--August 21, 2000 Leveraging the knowledge of industry experts from around the world, SPSS Inc. (Nasdaq:SPSS), NCR (NYSE:NCR) and DaimlerChrysler announced the completion and publication of the CRoss-Industry Standard Process for Data Mining (CRISP-DM). Released at the KDD-2000 Conference in Boston, MA, CRISP-DM 1.0 is a comprehensive data mining methodology and process model that provides anyone -- from novices to data mining experts -- with a complete blueprint for conducting a data mining project. Offering step-by-step direction, tasks and objectives for every stage of the data mining process, the CRISP-DM 1.0 user guide and reference document is available to the public at no charge. "The completion of CRISP-DM 1.0 represents a collaborative effort of top thinkers in data mining," said Colin Shearer, vice president, Data Mining Business Development at SPSS Inc. "Our consortium of industry experts, with input from approximately 200 representatives from other companies, worked together to create a standard, open and proven approach that anyone can use as a reference in their work. CRISP-DM is already accepted as the de facto standard for the industry and already, many organizations are recognizing this standard as the opportunity to increase the success of ROI." The CRISP-DM model makes large data mining projects faster, more efficient and less costly by enabling users to take advantage of an already-proven process. The model helps people avoid common mistakes and provides a head start in - Assessing business issues and articulating data mining methods to solve them - Capturing and understanding data - Identifying and solving problems within the data - Applying data mining techniques - Interpreting data mining results within the business context - Deploying and maintaining data mining results - Capturing and transferring expertise to ensure future projects benefit from experience SPSS Inc. joined NCR, DaimlerChrysler and Dutch insurer OHRA to create this model in January 1999, after acquiring consortium member Integral Solutions Limited (ISL). Since its beginnings in 1997, the CRISP-DM Special Interest Group (SIG) has also followed and supported the consortium's work on the model. The SIG has more than 200 members worldwide and includes large-scale data mining users from a wide span of commercial industries, as well as data mining software and service suppliers including IBM, SAS Institute and all Big 5 consulting firms. The broad range of experience involved ensures that the CRISP-DM model applies to businesses in any data mining application -- including fraud detection, credit risk, customer retention and churn analysis. SPSS Inc. uses CRISP-DM in delivering the professional services portion of its analytical solutions, including CustomerCentric, the company's flagship analytical customer relationship management solution. Many of NCR's Data Mining Laboratories, especially those in Europe, are using CRISP-DM to structure and define the data mining services that are given to customers and NCR's latest data mining solutions are written using CRISP-DM. ProofPlus, one of NCR's latest solutions for predicting churn, is based upon the CRISP-DM phase and task structure. Data mining plays a part in the success of many of today's businesses by helping to identify the hidden patterns in customer data that can improve customer relations, increase revenues and reduce costs. Once discovered, this knowledge enables companies to leverage their information and experience to make more accurate predictions and better decisions. The CRISP-DM initiative was partially funded by the European Commission, whose membership includes SPSS, NCR and DaimlerChrysler. Availability Visit www.crisp-dm.org to download a PDF file of the CRISP-DM document, incorporating the complete CRISP-DM 1.0 Reference Model and User Guide, free of charge. |
Copyright © 2000 KDnuggets. Subscribe to KDnuggets News!