Many business enterprises have accumulated massive databases containing valuable information about their customers, sales, competitors and products. In addition to the direct information captured in the database, there are gems of hidden information that cannot be easily accessed via simple query facilities.

Data mining is used to discover these hidden relationships and patterns in large databases. These specific correlations between entries will uncover associations within the database that may have been previously unnoticed. To discover these relationships, many hypotheses or data models must be examined by the computer, and an IBM SP can rapidly test thousands of data models in parallel. Deep Blue, a specially modified version of the SP, can examine and evaluate up to 300 million chess positions per second. This tremendous speed and capacity is quite useful when performing data mining tasks. In the blink of an eye, hypotheses can be verified and used to uncover the new hidden relationships useful to the business or enterprise.

Data mining is in use in many commercial businesses today, including weather forecasting, marketing analysis and even the National Basketball Association, where searching through millions of statistics and variables can help a coach put his most effective players in the game (the SP was utilized during last year's NBA playoffs to determine the optimal strategy against the Chicago Bulls). By applying the same power that drives Deep Blue - at one time only used in the scientific research community - we can tackle entirely new classes of business applications.

Other applications of Deep Blue technology

Financial modeling
Marketplace trends, risk analysis and more

Molecular dynamics
A valuable tool in discovering and devleoping new drugs