• OAL has adopted CRISP-DM (Cross Industry Standard Processing for Data Mining) as the proven methodology to build predictive analytics models across the enterprise. CRISP-DM is a non-proprietary, application neutral model that was developed to standardise the industry and provide companies with a roadmap to more successful data mining by encouraging best-practices.
• OAL has been involved with CRISP-DM since its inception, when founder & principal of OAL, Ralph Overbeck, was part of the community that brought CRISP-DM to life.
• CRISP-DM works because each phase of the process focuses on understanding the project objectives from a business perspective. Knowledge gained from data analysis can then be fed back into the organisation to aid business decision-making.
• The six phases of CRISP-DM are frequently referred to by analysts and journalists as the de-facto standard for the data mining industry. A poll of data mining practitioners, conducted by KDnuggets, revealed that CRISP-DM was the data mining process most often used by companies; 51% versus 12% using the second standard process - SEMMA

The following diagram illustrates the six phases of CRISP-DM.

Source: CRISP-DM web site.