RSS Filtering Now an API with PostRank
Canadian startup AideRSS is one of the coolest things since sliced bread. No, really. That may sound like hyperbole, but it is one of the most useful utilities I’ve come across over the past year. When I first read about developer Ilya Grigorik’s plans for an RSS filtering algorithm in December 2006, my first thought was, “I have to get this guy on the phone and convince him to turn this into a web service.” I procrastinated, but Grigorik didn’t and in July 2007 he launched AideRSS, an RSS filtering application based on his initial PostRank idea.
PostRank ranks RSS feed items by looking at how much user interaction they have had — the more user interaction, the higher the rank. Ranks are normalized to each feed, so that while 10 comments and 40 diggs on one blog may result in a PR of 10, it may only be a PR of 3 on a more popular web site. PostRank uses diggs, del.icio.us saves, Twitter tweets, Google Reader shares, comments, and starting today Ma.gnolia bookmarks, and Pownce messages to determine rankings.
AideRSS also announced today that it was spinning off PostRank as an API, available on the new PostRank web site. Opening up the PostRank algorithm as a service does two things: it means that anyone can now create “thematic” versions of AideRSS that mix and match articles on specific topics from different sources — you can apply rankings to posts from more than just a single feed now — as well as add post filtering to any application, and it may actually improve PostRank itself.
Because pageviews and article clicks are figured into the PostRank algorithm (right now via AideRSS widgets and RSS reader integration — such as via the Google Reader extension, which was updated today, or via NewsGator Online, which uses PostRank), any RSS reader or application that adds PostRank via the new API might theoretically send data back to AideRSS that could improve the algorithm itself.
As long as you can get on board with the idea that increased social engagement indicates better quality content, then AideRSS and PostRank are very helpful tools for cutting down the noise that any heavy media users feel. PostRank really shines because each post’s ranking is normalized against the average for that blog — thus rankings don’t favor heavily trafficked sites. Sites that have very little play on Digg or del.icio.us can still benefit from ranking posts with PostRank as a result.