Thursday, July 10, 2008

Data Mining news

It's been seven months since I barely pulled off my Master's Report and graduated. Over two years since I took my favorite class, which was Data Mining.

I have every intention of following up on the advice of my reader from the UT Journalism department, Dr. Iris Chyi (thanks again!) and finding some peer-reviewed journals to publish the study in. I'm told that it will take a lot of extra work first, which I haven't been sufficiently motivated to do. What I need to do is apparently read some journals that deal with the appropriate subject matter (i.e., focus on digital media for example) and get a feel for what kind of work they've accepted in the past. Then trim the paper to a manageable length (from 60 pages to, say, 25) and submit something that, in my judgment, they'd be likely to approve with no major changes.

I'm writing all this down so I have a reminder to actually get off my ass and do so, because otherwise I'm afraid I'll forget about it.

Determined to leverage this experience and make people remember who I am at work, on Tuesday I did an hour long presentation for most of the engineering staff (a room full of maybe 30 people) on what data mining can do for my company. I know I am an ultra-nerd, but I love doing presentations. False modesty aside, years of practicing topic presentation on TV and podcast have made me pretty good at it.

I put a Power Point presentation together and now I've uploaded it to Google Docs, so you can check it out by going here. You can't really get the point without hearing me talk, but it will give you a general idea of what I covered. The official topic title I picked was, "Data Mining: How math helps us compare apples to oranges, and shows that ice cream causes shark attacks." (If you care, I can explain that in the comments.)

Anyway, couldn't have asked for a much better result. I managed to hit all the major points I wanted regarding a complex topic in under 90 minutes with time to spare for questions. They laughed at all my jokes in the right places, appeared to follow the point of what I was saying, nodded sometimes, participated where I intended them to, and discussed it after it was over. All four of the company software architects had an animated discussion about what use they could make of it, after everyone else had left. And even today, two days later, I'm still getting IM's and people stopping by to let me know that they enjoyed and appreciated it.

People, please DO NOT FEED THE EGO!!!


  1. Anonymous5:07 PM

    The "ice cream causes shark attacks" bit is fairly common amonst anyone trying to explain the misuse of statistics. What's the phrase? "Correlation does not imply causation."

  2. Yep, that exact phrase is in the Power Point, if you'd take a look. :)

  3. Not feeding the ego here. Just a tip.

    The kind of reaction you got. And the bottom line results (or potential, at least) for the company are EXACTLY the kind of ammo you need when you brace the boss for a raise or promotion (or at your employee review...or, if the scat hits the fin, downsizing time).

    Follow up on it. Let your boss know what kind of feedback you're getting. Ask him or her for ideas on how it can be further developed/exploited for the company's benefit. Let it become, if only in part, her project too.

    Another benefit: just getting your fellow workers excited & motivated about something. Management often has a very hard time doing that.

    Just sayin'...