The UC Berkeley Entreprenuership Blog

Subscribe

A Review of Digg’s New “Upcoming Recommendation Engine”

June 30th, 2009 in Guest Posts by Aditya Mahesh
Digg Rec Engine

A few months ago Digg removed their “shouts” feature which allowed members to directly recommend stories to their friends on the website. This feature certainly made the social media site more social, but also led to an increase in spam as certain groups of users directly shares stories with each other and these member’s stories were promoted to the homepage of the site far more than those of other members.

Hence, Digg axed the shouts feature. Yet, a void remained. There had to be an easy way for users to share content with each other, but could not be so easily manipulated. After all that is the goal of any social media website. Digg recently implemented an upcoming recommendation feature which shows content submitted and dugg by members the engine determines “compatible” with the user. The way the engine works is a mystery, but it allows for users to get recommendations from other members, without allowing users to have direct control over this recommendation process. This eliminates the risk of users gaming the system, but places limits on how social the site truly is.

Still, most of the diggers I talked to said they liked the changes. For the most part, they said the recommendation engine did a good job of finding related stories and members with similar interests.

However, a major concern was the fact that there was no sense of control with what the engine provided. Digg power user FirstDigg told me that “the biggest problem (of the recommendation engine) is the untrainability & lack of scope.” He wanted Digg to give users more power over what stories appear allowing “trainability without actually burying a story. For example, a story can show up in the recommendation engine about team x winning the World Series. It’s news, sure, so it shouldn’t be buried, I just might not care about it and want to remove recommendations similar to it in the future.”

In my experience, the recommendations Digg is giving tend to be quite good; however, they are seemingly random. They are all on different topics, and while all stories are for the most part interesting, it either means I Digg a wide variety of story types or Digg is just throwing stories that other Diggers like at me. In addition, I find there are a lot of stories not in the recommendation engine that I am interested in so while I do check out what Digg recommends, most of the stories I like appear elsewhere on the website.

One thing I do use a lot is the “Diggers Like You” feature which uses some compatibility metric to show diggers who submit, digg, and comment on stories similar to those whom I digg, submit, or comment. I have used this feature to find a number of interesting members to befriend.

Share and Enjoy:
  • Digg
  • StumbleUpon
  • TwitThis
  • del.icio.us
  • Yahoo! Buzz
  • Techmeme
  • Reddit
  • Google
  • Facebook
  • Technorati
  • Mixx

One Comment

Omer

July 13th, 2009

Not a fan. The new diggbar is lame

Leave a Reply

Related Posts

Check out some more great tutorials and articles that you might like.