Data Mining Fundamentals

I love baseball. That being the state of things, I'm struggling to write this blog entry while simultaneously watching the Astros go at the Braves in the National League Division Series. I'm happy to report the Astros are winning -- not because I particularly like them, but because I can't stand the Braves. All this post-season baseball has got me thinking about fundamentals.Baseball players often spend too little time practicing the basics: base running, bunting, sliding, setting the feet before the throw. It's always worthwhile to go back to the fundamentals. The same is true in IT, and this week we take a look at the basics of data mining.

As data mining author Warren Thornthwaite writes in an article we excerpt from Intelligent Enterprise, the data mining process begins with an understanding of business opportunities. At the outset, identify and prioritize a list of opportunities that can have an impact on your business. It's that sort of data mining fundamental that helps an organization get a solid start with data mining. From there, it's on to building data models, followed by implementation, assessment and maintenance. Thornthwaite will take you through them all in detail.

We're also running a case study you should check out. Needless to say, it's not easy to manage 50 separate database systems. But Stride Rite has simplified the process by running graphical management tools on top of its Oracle 8i and 9i systems. The tools, from Quest, let Stride Rite's database manager pinpoint performance issues within the company's vast array of data sources.

And as always, turn to our Business Intelligence Pipeline Hands On section for more how-to's and reviews.