Instead of showing users the most popular stories, [Digg] would make guesses about what they'd like based on information mined from the giant demographic veins of social networks. This approach would essentially turn every user into a big Venn diagram of interests, and send them stories to match.Please see also my older posts, "Digg recommendation engine" and "Combating web spam with personalization".
Adelson said Digg had not yet deployed local views of the content, but that it was in the planning stages. "We do believe the implicit groupings of users and interests that we use in the recommendation engine will certainly play a role in the future of Digg and how we can address localities and topics."
Update: It took a while but, a year and a half later, Digg is about to launch a new version of their site focused on personalized news.
Update: Two years later, finally, Digg launches its new website. As Mashable asks, too little too late?