Benford's Law Meets Kirix Strata

Posted July 24th, 2008 by Taran

Back in February of 2005, I headed up to Boston for Linux World as the Editor of LinuxGazette.com - I looked a tad different - and one of the last parts of the show I got to see was Kirix Strata. I wrote about it on the now gone LinuxGazette.com site, as I recall, and I remember seeing Kirix Strata as a solution waiting for problems. I had a licensed copy, but never seemed to have the right problems to toss at it.

Ken Kaczmarek, who I met at the show, dropped me a note about a post from the Kirix blog which he thought I would find interesting: it deals with the all too often forgotten Benford's Law. In a nutshell, Benford's Law states that in lists of numbers from many non-artificial sources of data, the leading digit is distributed in a specific, non-uniform way. Take a look at the video for a better understanding.

Pretty Cool On A Few Different Levels

Aside from the ease of use of Kirix Strata demonstrated in conjunction with Benford's Law, I was impressed with a few other things. First, I hadn't been paying attention to Kirix because I've been busy with other things - and I didn't even know that they had a blog. Next, the whole demonstration of Benford's Law is probably the best true piece of marketing I have seen - not because it will make you want to buy the software, but because it demonstrates Kirix Strata with some real data in a context that gives Kirix Strata a voice. I'm not a fan of most methods of marketing, but this may well educate a few people in some cool math while showing a bit of Strata's muscle without it ripping it's shirt off. The video uses data from the Census Bureau, the Wikipedia and Digg.com itself.

Yes, I did download Kirix Strata's 30 day free download - but doing the things that I am doing now (agriculture/land management/etc), I have actual uses for it that I'll be playing with now that I have a group of problems to throw at Kirix Strata. I'll post some of my findings as well... And there is an open source twist to this product that I'll cover sometime tomorrow (I have to revisit a few things so I am sure of what I write).

If you want to read more about Kirix Strata, you should check out the overview. While it is geared toward research analysts, web developers and business analysts, the power of Strata is worth exploring for anyone whose data problems justify $250 US. I actually do believe I have enough data on some things to make it worthwhile myself.