Saturday, as I was writing a blog post about Building a Community around Football and Food on Bainbridge Island after my son's first football game, I noticed that the News pane of my beloved Google Desktop Sidebar started showing me sports headlines. In a short space of time, I learned of a number of developments in which I have absolutely no interest, e.g.,
Having a background in information extraction and personalization, I've often wondered what kind of algorithm Google uses for both creating a profile of my interests, and how and where it searches for articles to show me in the Sidebar that are likely to be relevant to those inferred interests. In writing the blog post, I had visited the home page for Evan's football team, the Woodinville Falcons, as well as the Bainbridge Island football team. But those are the only sports-related pages I've visited since last summer, when we were given free tickets to see the Seattle Mariners. So I was surprised when articles about sports -- and, more specifically (from what I can gather), football -- started appearing in the News pane of the Sidebar on Sunday, with a frequency of once every two or three articles.
So, I started wondering about what might explain this high frequency of football-related pages being shown, given the relatively small sample of football-related pages I visited. Could it be that terms relating to football were so prominent on the two pages I visited that those terms became highly rated in my profile? Could it be that there were so many articles about football in the pool that Google draws from (especially on the first Sunday of the NFL season), that even a slight hint of interest in football results in a flood of such articles? Or, could it be that my Google Sidebar has gained some special insight into my true character -- that beneath this facade of disinterest in all professional sports, I am, deep down, a football fan?
I'm reminded of an article in the Wall Street Journal a few years ago, "My TiVo Thinks I'm Gay", in which a number of people discovered that their TiVo digital videorecorders seemed to be making unwelcome inferences about their interests, and tried to "correct" their profiles:
... when TiVo thinks it has you pegged, there's just one way to change its "mind": outfox it.
Mr. Iwanyk, 32 years old, first suspected that his TiVo thought he was gay, since it inexplicably kept recording programs with gay themes. A film studio executive in Los Angeles and the self-described "straightest guy on earth," he tried to tame TiVo's gay fixation by recording war movies and other "guy stuff."
"The problem was, I overcompensated," he says. "It started giving me documentaries on Joseph Goebbels and Adolf Eichmann. It stopped thinking I was gay and decided I was a crazy guy reminiscing about the Third Reich."
TiVo does provide more direct, manual intervention, offering users the capability to vote -- thumbs up or thumbs down -- on any of the television shows in its listings, but even that can lead to mistaken inferences:
Mr. Karlsson, 26, says he "pre-emptively" found all the religious shows in his TV listings and used the "thumbs down" button on his remote control to tell TiVo he has no interest in them. (Giving three thumbs down is the best way to block a program.) After that, his TiVo recorded movies about creepy homicides. "They all have titles like 'Murder on Skeleton Isle,' " says the computer system administrator in Cambridge, Mass.
Google Desktop Sidebar also offers users an opportunity to directly intervene, but only to give a thumbs down via right-clicking on a headline and selecting "Don't show me items like this".
This helps, but I still want more control. When I give a thumbs down on the three articles I listed above, I don't want Google to adjust my profile so that articles involving chiefs, colleges or injuries no longer appear in my News pane, but simply filter out football articles. After telling Google not to show me more articles like the three listed above, plus about a half dozen others, it did stop showing me sports related articles, but now I'm wondering what I'm missing.
One of the nice features of the Sidebar is that I can continuously partially attend to it, rather than give it my full attention, as I tend to do with television during the one or two times per month I watch a show (which, of course, is nearly always recorded on my Dish Network DVR ... although with the TiVo patent lawsuit advancing, this practice may change). So the bar is lower, so to speak ... but perhaps not for long. With recent developments, such as YouTube and Google Video, we are seeing a convergence that is closing the gaps between television and the web, and it will be interesting to see how personalization and profiling will adapt to accommodate the growing volume of digital content.
[thanks to SnarkAttack for posting an excerpt of the WSJ article online]
[Update: I wrote another post yesterday on Football and Community in Woodinville. This time, I didn't even visit any other sports-related web sites (e.g., for the Bainbridge Island or Woodinville football leagues, as I did for the first post). And yet today, 7 of the 20 News items that Google Sidebar is showing me are sports-related (as shown in the screenshot on the left). This suggests (to me) that it is the preponderance of sports-related articles in the news pool today, moreso than a preponderance of sports-related pages I'm visiting ... unless my Sidebar still thinks I'm a sports fan (on Sundays).]