Hyperion and SPSS Partner
The announcement of a partnership between online analytical processing (OLAP) specialist Hyperion Solutions Corp. (
www.hyperion.com) and data mining stalwart SPSS Inc. (
www.spss.com) sets the stage for the convergence of OLAP and data mining technologies. According to some industry watchers, it's a combination long in the making.
In June 1999, Oracle Corp. (www.oracle.com) purchased Thinking Machines Corp. as a prelude to subsuming the latter's data mining technology within its own portfolio of business intelligence and data-warehousing solutions. Oracle has thus far unveiled the data mining tool dubbed Darwin, which is based on the Thinking Machines' technology. The company hasn't, however, announced an integrated OLAP and data mining strategy.
The Hyperion/SPSS partnership is grounded in this combined strategy, and requires only execution on the part of its two member companies, says Mike Schiff, director of data-warehousing strategies at Current Analysis Inc. (www.currentanalysis.com).
According to the terms of the partnership, SPSS will integrate its DecisionTime and WhatIf? forecasting and predictive analysis products with Hyperion's Essbase OLAP server. Both companies say the combination of the two technologies will allow customers to use Essbase's OLAP capabilities as a means to derive enhanced business metrics from the forecasts and predictions created by SPSS' data mining tools.
"The OLAP structure provides one form of doing analysis, but with the addition of data mining technology, you can enhance the value of what's contained within the OLAP structure itself by forecasting, by predicting," says Douglas Dow, vice president of business development at SPSS. "With the forecast, you can put forecasted values back into the cube and really enhance the value of what's there by adding this predictive element."
According to Don MacTavish, director of strategic marketing at Hyperion, the combination of OLAP and data mining technologies can take some of the black magic out of business forecasting and out of business intelligence in general.
"With the combination of the SPSS technologies for doing very powerful forecasting and our OLAP technologies, when you mix the two it really will produce some very powerful results," MacTavish claims. "For many people when they say forecasting, they'll sit around a room and guess numbers and ultimately hope to grow by 10 percent, but what we're doing is combining a predictive technology with our OLAP sever that mathematically calculates a forecast using a series of models."
Because OLAP engines can provide verification services for the discoveries, forecasts, and predictions unearthed by data mining tools, Current Analysis' Schiff says, marrying the two together is sound strategy.
"What this does is it lets you develop predictive forecasting models and deploy them throughout your organization," Schiff explains. "I've always been an advocate of data mining and see it as a logical extension to the analytical capabilities of OLAP servers in general, so the combination of SPSS' technology with Hyperion certainly makes a great deal of sense."
Moreover, Schiff suggests, the Hyperion/SPSS partnership provides an indication of the growing acceptance and importance of data mining across the industry as a whole.
"Oracle bought [data mining specialist] Thinking Machines, so how they implement that technology still remains to be seen. But certainly data mining is becoming more and more of a mainstream technology these days," he concludes.