Back in the day, Google’s pagerank used many individual acts of linking to calculate the relevance of pages. These days, the acts of linking are occurring in near-real time and viral waves on Twitter and social network services. Links in social media are a powerful indicator of relevance. A link has been retweeted, friendfeeded, bookmarked and facebooked. So there could be a “social network page rank” algorithm that calculated the relevance of links, and a “network digg” that showed the heat of a meme.
One problem is that twitter’s 140 character limit encourages the use of url-shortening services like bit.ly and tinyurl, which obscure the destination and content of the link. So the service would need to expand and compare the urls, and do a little analysis to figure out when slightly variant urls link to the same content.
A secondary problem is potential spam, but a social white list – in Clay Shirky’s geek-felicitous term, foaf-filtering – could mitigate that. Social self-promotion (I’ll retweet yours if you retweet mine) could be a problem, but I suspect the echo chamber effects would be fairly localized for garden variety topics, and the pop culture or political fangames would be interesting in their own right.
Does this exist yet? Urls welcome.
Update 2. John Battelle The Conversation is Shifting on the trend toward social search. I don’t think it stops at neophilia.