IBM announced a standards-based data mining service Nov. 2 that it says makes it easier for businesses to develop targeted, personalized relationships with customers and suppliers.
The service, called the IBM DB2 Intelligent Miner Scoring Services, is a data mining application that allows companies to rank customers according to predetermined criteria. This allows a telemarketer at a bank, for example, to determine a customer's creditworthiness in real time based on the institution's policies while an applicant is on the phone, or even automate it as part of a self-service Internet process.
The new scoring services will be sold as an option with DB2 Universal Database and, as a result, will directly integrate data mining technology into the relational database management system for rapid application development and deployment and faster application performance. This integration allows mining models to be updated dynamically, so there can be a constant improvement in business results without disrupting production systems.
At the heart of IBM's data mining capabilities is the industry standard for predictive modeling, Predictive Modeling Markup Language. The Data Mining Group, www.dmg.org, working with representatives of the data mining software industry and from academia, defined this standard to easily create and share data mining predictive models. PMML is an XML-based language that provides applications with a vendor-independent method of defining predictive models so proprietary incompatibilities are no longer a barrier to the exchange of models between applications. Predictive models express patterns of information discovered in data mining.
DB2 Intelligent Miner Scoring Services is in a beta test program with early customers and partners. The services are expected to be generally available early next year.