Greg Linden, founder of Findory.com which (I think) is the best news and blog personalization out there, has blogged that Google just released recommendations on Google News:

This would suggest that the underlying recommendation engine either is using some combination of subjects and similar subjects (e.g. read an article on internet technology, get recommendations of popular articles in subjects similar to internet technology) or a clustering approach (build a large number of groups of different readers with various interests, match you to the most appropriate cluster based on your history, recommend things popular with other readers in the cluster).

That's okay, but the problem is that the recommendations will tilt toward the popular and away from the tail. You want the recommendations to be surprising, to enhance discovery, to help me find things I wouldn't have found on my own. Clustering or subject-based techniques won't get you there.

How is this related to social networking? Well, this may be a stretch, but it's not totally off subject. Unlike Greg's findings and theorizing, Google says they are using collaborative filtering, which would imply that other people (unbeknownst to me) are affecting what news I see - based on what they click on:

Google News can suggest news stories just for you. If you have Personalized Search enabled, you can sign in to your Google Account to get recommended news stories based on your past news selections.

Google News can ... compare your tastes to the aggregate tastes of other groups of similar Google News users. Simply put, we recommend news stories to you that have been read by many other users who've also read similar stories as you in the past.

We're seeing a rise of two different types of collaborative filtering: algorithmic and explicit voting of social networks. Whereas Digg, reddit (and deli.icio.us to a degree) are following the explicit vote path, findory and google news are following the algorithmic path. It'll be interesting to see how these two approaches merge.

Why do I think they will merge? I think that both approaches will eventually fall prey to spam. It is obvious how the explicit vote sites can be succumbed by spam. But, algorithmically personalized sites are subject to click fraud methods as well. Stan, from Outfoxed, spells out a similar scenario:

Alex suggests that the best way to protect the beer is with more complicated algorithms, just as Google revived search after the exploitation of early naive search engines. Here I have to disagree. The answer will inevitably be to use only trusted sources, and to give up trying to figure out who the bad guys are. Just let the people choose their own sources, recurse outwards through social networks, let small world networks do their thing, and *voila* you've got good data.