Social networking site Digg has a beta going for their Recommendation Engine, which suggests links that users with similar interests have dugg:

Every time you Digg a story, the Engine matches you with other Diggers who Dugg the same story, and keeps track of all your Diggs in common with them.

When it’s time to calculate your Recommendations, the Engine draws from this pool of matched Diggers. … [Y]our Recommendations consist of stories that your Diggers Like You have already Dugg, minus the stories you already Dugg or Buried[, from topics] where you have a shared Digging history. We figure that two users may have similar interests in a subject like “playable web games,” but one person might be into politics while the other follows celebrity gossip. So we actually compute correlations, Diggers Like You, and compute Recommendations in several collections of topics independently.

It’s a good idea in that Digg has become overrun with uninteresting articles. I joined the site in May 2006 because it pointed me to sites I found worthwhile; I filtered out sports and music news and waded through the rest. Today, a large percentage of the links are spam, politically biased, or childish YouTube videos — and my filter list has expanded to include all videos, images, and political sites.

Having the Recommendation Engine will certainly cut down on the number of bad links I see, and undoubtedly get more participation from lazy users who, like me, read articles without giving a thumbs up or down at the end. Still, with it you lose some of the obscure treasures that defy categorization.

Sadly, Digg seems to have unwittingly released a free targeted ad system upon its visitors. While I trust my friends not to recommend bad sites, I have no trust in the kindness of strangers to recommend good sites. If LG wants to advertise their latest iPhone knockoff, they can just create a user and go to town Digging all stories relating to touchscreen cellphones, then plant their own press release. Without the Recommendation Engine, their link would undoubtedly wash away in the torrent of submissions, but now it is prominently targeted to the users who are most likely to be interested in that product.

Even worse, I imagine you could game the system even more by “stalking” a user, mimicking their actions (as seen in their History) to create a 1.0 Jaccard index in each topic. Combine this with the Who Dugg It? feature, and you could find users who showed interest in your product or a competitors and advertise directly to them.

I expect Digg will have to address the overwhelming potential for abuse before the Recommendation Engine leaves beta. I’ll be watching it with interest all the same — there’s nothing else to do on Digg.

This entry was posted Monday, July 7th, 2008 at 11:08 pm and is filed under Computer Science, Secrets, Security. You can leave a response, or trackback from your own site.