Adotas

Where media buyers start online

Featured Author

Author Photo

More articles by Kenneth Musante





Features

Identifying Social News Clustering

Written on
Sep 8, 2006 
Author
Kenneth Musante  |
Share
Identifying Social News Clustering

There’s an interesting phenomenon happening on Digg, the ever-growing social news site. It’s something that appears to be part of a natural progression of the site’s open-ended and democratic editing procedure.

The site functions as it always has: users submit links to news stories, and vote to determine which links get promoted to the front page. Each person gets one vote. Digg founder Kevin Rose has always stated that Digg takes a democratic approach to news. In a homogeneous world, the existing social algorithms the service uses may produce just that. But just like the real world, Digg, as it grows, is experiencing a little bit of clustering.

Because of the way the site works, groups of users working together, whether deliberately or inadvertently, have a greater influence over what does and doesn’t appear on the Digg home page. This is especially true in categories where user opinions have polarized (like Left vs. Right in politics). I’ve also seen some users complain that Scientologists are “burying” anti-Scientology stories.

This sort of user-clustering is something that the infamous Jason Calacanis trumpeted loudly about after he launched the new Netscape, which has much the same functionality as Digg. And while some people roll their eyes whenever Jason opens his mouth, and I have no way to verify his numbers (I’m not even sure where he got them in the first place), the gist of what he said is true” “30% of the front page stories come from the top 10 users on Digg.”

Digg’s algorithm for promoting links to the front page has always remained secret to protect it from those who would deliberately manipulate it for their own gain. It takes into account spam-like behavior to render digs from users with multiple accounts or machine-like digging patterns impotent.

But clustering is a legitimate result of how Digg and other social sites work. Every social site contains clusters of users huddled around subjects of common interests. A site dedicated to something specific like a forum dedicated to Halo 2 may contain only one cluster, while a social blog network like Livejournal contains many. The number of clusters increases as the number of users increases and the scope of the site broadens.

And with a site like Digg where the users are the sole determiners of what’s hot, clustering produces some users that have greater influence over it than others. The launch of Digg 3.0 saw the addition of mainstream news subjects like politics, sports, world and entertainment, opening it up to more than just tech users.

Kevin’s idea for Digg has always been a utopian one. No editors. All users are registered as equal, with certain inalienable rights endowed by Digg. But as Mike Arrington of TechCrunch says, “If Digg is being corrupted by a relatively small group of users, the difference between Digg and the NYT becomes less clear.”

To address the anti-democratic results of clustering, Kevin announced that he and his team will be updating the Digg promotion algorithm to take away some of the power held by cluster leaders (”users that follow a gaming pattern”) (gaming=cheating). The changes will also give greater weight to users who consistently digg stories submitted by lots of different people, rather than just a few friends.

It’s a good move, and it holds fast to Kevin’s idea of a completely democratic news site. But it is a sharp blow to some of the top users who digg and submit stories with an incomprehensible passion, and for whom maintaining a top Digg ranking is a heated competition.

When Kevin posted his announcement on Digg, it was immediately met with a flurry of comments from users, both for and against. It was Digg’s unpronounceable top user, p9s50W5k4GUD2c6, who effectively quit Digg and expressed frustration at having worked so hard to maintain the top Digg spot, only to have the algorithm that made his status possible pulled out from under him by commenting “I will NOT overlook your tacit equation of BUSTING MY A** for Digg with gaming.”

While users dictate what appears on Digg, they’re not dictating Kevin’s vision for the site. It shows internal strength and a high ideology. Kudos to Kevin. I’m sure Calacanis’s Netscape will face similar problems if it doesn’t crash and burn. I don’t think there’s any way to eliminate clustering. In fact, I think clustering is vital to human behavior and social interaction. Something that’s an important part of the social Web. But it’ll be interesting to see, when Digg starts using the new algorithm, if there’s a mathematical way to mitigate its effects.





Reader Comments.

I supported thoughtful changes to the Digg algorithm that will broaden participation on Digg. As long as it is implemented correctly, it should prove helpful.

While it was not Kevin’s intent, I left Digg because top submitters were associated with gaming the system. I found that offensive. My departure had nothing to do with changes to the Digg algorithm.

Respectfully,
P9 (AKA: p9s50W5k4GUD2c6)

Posted by P9 | 1:02 am on September 9, 2006.

Leave a Comment

Add a comment

Tags: and
Article Sponsor

More Features


  • Are you taking advantage of Twitter's new geolocation service?
    View Results

    Loading ... Loading ...

Polls Archive

Latest News

News Archive

Spotlight

AdBidCentral’s CEO, Vivek Veeraraghavan Talks Openly*What was the inspiration to start AdBidCentral? The conditions that inspired AdBidCentral came from a variety of factors in my personal [...] more...

Adotas Partnership