Clay Shirky, in this recent session at Web 2.0, talks about how the filter has moved downstream from those who are responsible for publishing, to those who consume. The initial, visceral, effect of this is the well-known and overhyped “information overload” meme. In the above talk Shirky places the blame for information overload squarely on the head of outdated filters, a situation he labels “filter failure.” Below is a step in the right direction.
In this very interesting post over at ReadWriteWeb about online noise, Marshall Kirkpatrick mentions a wonderful experiment in noise control: AideRSS. The service uses a five-element filter called PostRank to dynamically measure an engagement facet. This works by measuring the engagement resulting from individual posts on a scale from 1 – 10 (with 10 being the most engaging). The theory is that the more a post stimulates engagement, the better the post. PostRank uses what it calls the “5 C’s” to measure engagement:
- Creating: a post that causes others to create something original
- Critiquing: leaving a comment on the post
- Chatting: sharing the post via Twitter, etc.
- Collecting: sharing via a social bookmarking service
- Clicking: simple page views
The result is an enhanced method of filtering large amounts of postings, like this example of AideRSS in action on Firefox using Google Reader:
Filtering Multiple RSS Feeds at a PR score of 6.0 and Over
What’s interesting about filtering multiple feeds is how useful it is. While Google Reader reports 363 posts for my Web2.0 folder (which currently holds 18 feeds), filtering for the “best” knocks that list down to a mere 11 postings. 363 is more than I can really digest, and when it gets to be that heavy I am sorely tempted to “mark all as read” so I can resolve to keep up later. AideRSS allows me to grab just the most active postings and forget the rest if I don’t have time. (and who does, really?)
Further, the AideRSS filter allows you to analyze a specific RSS feed, like this recent feed from ACRLog:
ACRLog with filter set to 'all'
This gives me a recent history of postings from which I can get a sense of the quality of the feed. I can even add up the scores for 100 postings and then average them (which I won’t do, not enough time.) But that would give me another view of the feed, kind of like an impact factor for weblogs.
This is an exciting tool, with lots of potential for libraries if we could only find a way to capture an engagement facet.
3 Comments
You may want to take a look at Paul Pival’s post on odd results he had using AideRSS.
Thanks Stephen. I think Paul’s right, they do need to work on their algorithms for PostRank. I have noticed that the age of a post determines whether it shows up in the “best”, etc. filter. But their scale seems oddly implemented. This post itself will probably shoot up in PR even though it’s a low traffic blog but with 2 comments. It may be really easy to game this system.
Hi Steve — Great write-up, thank you!
Also, thanks for the heads up on the Shirky video. I’d seen it linked but kept forgetting to view it. Juicy food for thought on how we’re evolving (and occasionally stumbling) in reinventing our social models.
I’m totally on board with the idea that information overload is just “what we swim in”, and, as with many things, the upcoming generation will accept and manage that better than those of us who are still trying to get used to it — and still having emotional reactions to it. (Honestly, how odd is it that we can experience guilt over data?)
Of course, those who’ll manage information overload as a matter of course won’t have the same filter failure, since they’ll accept information management tools as a matter of course. Largely, I think, because many of those tools will be seamlessly integrated into their lifestreams.
RSS is a great example of this. Far more people use it, and will use it, than understand or recognize it. In our little corner of the world, ideally it would mean an audience who never knew that RSS and many other tools never existed without analysis and filtering built in.
Post a Comment