Note that Tivo doesn't say if you love CNN, MSNBC, and the
History Channel, you'll love the Discovery Channel. The
recommendations are at the individual show level. Now
that's what I'd like to see.
A spambayes in
reverse. Read a few blog entries and push on little green
thumbs up or red thumbs down icons. Then any all all posts on
topics that interest you in your greater
neighboorhood will come to your aggregator.
You mean yours still forces you to enter URLs manually? I didn't think there were still such dinosaurs in existence. Next you'll be telling me it doesn't even support RSS autodiscovery.
Then again, how long will it be until we start seeing articles in the Wall Street Journal: "My news aggregator thinks I'm a pregnant gay man on crystal meth"?
For the record, I started auto-recording Futurama and Looney Tunes, and suddening my TiVo started suggesting every cartoon under the sun, including Blues Clues and the Powerpuff Girls. Apparently there are 4-year-old girls running around watching Futurama. I fear for the next generation.
Mark,
Apparently your TiVo is getting it's recommendations from my TiVo, as PPG, Blue's Clues and Futurama are regular shows at our house. Have no fear though, the kids aren't watching Futurama, yet...
Makes me wonder if TiVo takes geographic location into account when doing recommendations.
Comment spamming will be simple with a standardized API. :(
On the recommended front, my aggregator keeps track of which sites I click through to, and then sorts my feeds by the clickthroughs. That way the sites I visit most end up at the top of the page. I also want to add the thumbs up / thumbs down options, but I haven't implemented that yet. My first task right now is to get mombo working on my server, so I can switch over from blosxom.
Mark Pilgrim says that his aggregator is already Tivo-like. Mark's recommended reading page is pretty cool, but the algorithm recommended quite a few clunkers. My recommendations included the old blogs for Sam Gentile, Drew Marsh, Patrick Logan, and...
Gordon's point is valid: the blogging ecosystem data that "Recommended Reading" is based on is entirely machine-generated and machine-massaged. Lots of people are still linking to the old versions of weblogs (ask Sam about his RSS subscriptions sometime), so the ecosystem picks up on those links and my script recommends them. Garbage in, garbage out.
As for A-listers, well, the script tries to compensate for that somewhat (giving more weight to midlist blogs with fewer outgoing links than A-list blogs with tons of outgoing links), but power laws are power laws. I've tried about 7 completely different methods and algorithms for recommendations, and every one of them says I should be reading Scripting News. That was the #1 reason I implemented the "not interested" feature.
Prime example of Mark's point: w3future. This pages is dated March 30th, and look at the list of weblogs that he reads. The page he links to for me has not been updated for over three months.