Tailrank: Personalised "Google News" for Blogs
Tailrank is a memetracker, like memeorandum, and it groups related items together in a thread, like Google News. And naturally, threads with more buzz ranks on top.
If it had just stopped there, it would've just been yet-another. However, Kevin had the smarts to introduce personalisation. What does meme + personalisation give me?
Meme/clustering: soften the noise, group related items for me - so I can skim through faster, notice the big (arguably important) news and hence keep in touch with more sources (I'm currently reading ~200 feeds, people like Scoble reads over a thousand... !)
Personalised ranking: the idea of using OPML to help me personalised is quite ingenius. Now I don't have to explicitly tell the system what I like using hard, cold, insensitive keywords like, e.g. rss, ruby, ajax, etc.. I just give the system an OPML file (the defacto representation of a "subscription list"), meaning "here's what I read, show me what is important to me".. akin to Amazon's "here are the books I bought, now recommend me a new book"
Though I like the implicitness of importing OPML to "set" my preferences, the fact that an OPML file has to be used might probably leave out a larger population - if RSS adoption is 4%... OPML is probably...
I wonder if there're more straightforward sources could be used to infer... and hence making the service more inclusive? My del.ici.ous bookmarks? My flickr photos? My homepage? My browser bookmarks? My browser history (oh.. too private)?