Posted by Jim at December 5th, 2004

I’ve been interested in data mining for a while now. Basically, data mining uses statistical techniques to see patterns in the data you’ve collected in your data warehouse. Of course, you don’t just use statistics. You can also use artificial neural networks and other interesting techniques.

Who uses data mining? Large organizations, mostly. Hospitals use it to understand what treatments have the best effect (for the least money?). Large companies use it to predict what products they need in stock (if we get an order for product X, we get an order for product Y 5 weeks later). The federal government is probably trying to use it to predict who is likely to become a terrorist…

It’s an interesting topic. As such, it’s inevitable that there’s an open source based program to do it. Orange uses python, C++ and QT to create a graphical program for analyzing statistics. I like it a lot and hope to find a way to use it.