Forming a memetracker community
I was traveling on Saturday and driving all over Mountain View on Sunday collecting party supplies, so I completely missed the Battle of the Memetrackers challenge that Steve Rubel staged over the weekend. I don't agree with his methodology, but that is beside the point. What was more interesting was the dialog engaged in by Gabe Rivera of Memeorandum and Kevin Burton of TailRank in the comments of Steve's first post on this challenge and a later one as well.
Gabe: "The software deliberately omits links to posts ("memes"?) that it discovers, in order to improve readability. The idea is that not every post that links to X deserves to be read as commentary on X."
Kevin: "Gabe. So are you saying you indexed the posts and discarded the links? Memeo misses posts which I personally found relevant and wanted to follow."
Gabe: "Kevin: yes, indeed. That's not to say I index everything, but not all posts indexed are shown. For instance, the "links for 2006-02-19" type posts don't make for good reading and are routinely filtered."
Kevin: "Gabe. We ignore posts of the type "links for 2006-02-19" as well. I consider most of these invalid though. What I was mostly talking about was legit posts. TailRank picked up links which were legit which you didn't add to your cluster."
Personally, I find this stuff fascinating, which may say a lot about me. I'd love to see this type of discussion institutionalized in some way, and hopefully have a few of the other memetrackers brought into the discussion. What I'd like to propose is the creation of a group blog on which Gabe, Kevin, and a few other memetracker creators could discuss their philosophy of what a site of this type should do. There wouldn't be any discussion of algorithms and "secret sauce," but surely issues like whether a thread should be removed after 24 hours is worthy of discussion. (Gabe says yes on this, Kevin says no.) There is no right answer, but it would be great to see the motivations of these authors. The posts could be at any frequency, and hopefully the colloborative effort would alleviate the defensiveness Gabe sometimes demonstrates and the combativeness Kevin seems prone to. Comments would be turned on for this site, which would be where the rest of the blogosphere would have their say.
The benefits to the readers of these sites would be better products, and hopefully the creators of the sites would find the same benefit. Another benefit to the site creators would be a better understanding by the public of their individual philosophies. I'm actually more interested in the benefits to creators of new memetracker sites. It should be clear that this genre is in its infancy, and I can only imagine what would happen is some bright coder were able to observe this discussion over a decent period of time, internalize the issues, and then build on that understanding. Everyone would benefit.
We need an advisory board to help arbitrate the selection of memetracker participants on this blog. John Musser of ProgrammableWeb, and Dan Gillmor of the Center for Citizen Media have agreed to join me in this effort, and I think we may need one or two more. For the sake of fairness, this group should be replaced periodically.
Appropriately enough, the discussion of this idea will probably appear on the memetracker sites.
