Thursday, October 21, 2010

Stand & Be Counted: Data Miners Weigh In On the Long-Form Census Debate

I had high hopes for the Data Mining Forum on Wednesday, and I wasn't disappointed. I was absolutely fascinated by Stephane McGee and Obaid Rahman's presentation on how statistics are used in the insurance industry: incredibly appropriate fare for World Statistics Day! The 100+ attendees were buzzing with questions as we pushed deeper and deeper into break time. It's always great to see such interest generated out of a presentation: great job, Stephane & Obaid!

Victor Zurkowski then told a story of how he came to be aware of the possibilities of net lift and return optimization. A slow smile spread across my face as a slide of the SAS Tech Report was splashed across the screen. Edited by Waynette Tubbs (I highly recommend following her on Twitter, by the way: she always seems to have interesting things to say) it's full of SAS tips & tricks, news and developments in the SAS world and a whole host of user support options to boot! In no small way is our own insights newsletter a reflection of this resource. It was a mention of net lift in Waynette's editorial of an issue of the Tech Report which generated his interest in maximizing net lift and return, and he applied the concept in a fascinating way... you can find out more once his presentation is posted!

I'd been looking forward with baited breath for the panel discussion. As an admitted data junkie, I've been a little too obsessed with this topic since it came to the media's attention in June. In fact, the panelists and executive committee of the Toronto Data Mining Forum could easily attest to this by sharing the four pages of background resources I provided to help prepare for the discussion. A little overkill, perhaps.

At any rate, the Managing Director of Redwood Data Sciences, Mr. Chris Osborne, skillfully directed questions around the loss of long-form census data to members of the direct marketing, financial services, health information and consulting verticals. We hope to make some of the highlights of the discussion available soon, so stay tuned!

The presentations are being cleaned up a little as we speak and should be posted sometime next week on the Toronto Data Mining Forum website. If you can't wait, I've posted my social media presentation on SlideShare within LinkedIn: feel free to have a look, I welcome your feedback.

Next up for me: a week-long trip to the west. Calgary and Edmonton, I'm incoming! To get some insight into our guest speaker - the legendary Marje Fecht - feel free to check out this video interview I conducted with her a few weeks ago.

Thanks again to the fantastic executive team of the Toronto Data Mining Forum. They are truly emblematic of the sense of community and support we've been striving to build between groups of data practitioners. Without their help, the presentations and panel discussion would not have been as compelling. Thanks once again, team!

Until the next time, I'd like to present the same question here that I posed to the Data Mining LinkedIn group. Will your organization be impacted by the potential loss of demographic data caused by making the long-form census voluntary? If so, how? Let's not limit it to your own organization either: what do you think some of the ramifications might be?

I look forward to your thoughts!

1 comment: