18.9.07

You know what we need?

A Last.fm for movies.

OK, fine, I just realized IMDb has a recommendation system based on user reviews. But that's not it. I'm talking about a similar system that tracks users viewing choices and then creates most-likely to like lists based on data. It would suggest not just movies you haven't seen from people who's movie tastes most closely match your own, but also suggest buddies that are statistically close to you in taste, so you can go ahead and browse their stash directly.

If you're not already familiar with Last.fm, you can check it all out at... Last.fm. The crux of it comes from a little app you download that 'listens in' to your iTunes choices. And, you can friend me too, for that matter.

OK, now someone go do that. 50-50, right?

**Oh wait, Netflicks already does this.
(Thanks, Peter) And no I didn't know--- wow, but I like!

3 Comments:

Peter Dolan said...

Except for the actual suggestion of people who match your taste, that's exactly what has powered Netflix's success. A recommendation engine that tells you what you'll probably like (not just what you'll like, but how much you'll like it), and lets you follow the queues and viewed movies of your friends.

Also fun that they threw out a huge set of sample data for us statistics geeks to masturbate with, but I'd be surprised if you didn't already know all this.

melanie said...

you know, I think i do remember reading that NF released the data last year...

But then, what have people done with the data? -- And is it continually updated?

I don't use NF yet, but if you're registered is there an internal community surrounding the recommendation engine, like with Last.fm?

Peter Dolan said...

That's where we come in. Time to free the data.