More On Social Networks And Recommendations

Michael Giberson

Chris Anderson at the Long Tail says there is a problem with using social networks to generate recommendations:

The problem with social software as a recommendation network has its roots in the problem of social software itself. “Friend” is a pretty blunt instrument when it comes to describing relationships, especially in matters of taste. The sad reality is that most of my friends have rotten taste in music (I don’t hold it against them), while the music recommendations I actually follow are mostly from people I’ve never met, be they Rhapsody editors or MP3 blogs. Same for virtually every other narrow category where I need advice; odds are that the real subject matter experts aren’t anyone I know.

This isn’t exactly on point. Here Anderson is talking about the limitations of using an existing social network program like Friendster to gin up recommendations, whereas the Hargittai post at Crooked Timber that Lynne cites is directed more at recommendation engines build up on social networking principles.

In the comments to the Hargittai post, someone recommended the movie video recommendation website What to Rent?, that doesn’t appear to rely on social network models. Instead, What to Rent? uses a 20-question personality profile and an indication of what kind of movie you’re in the mood for.

I tried it out, and the first 5 or 6 recommendations looked reasonable. Some of the recommendations I had already seen (i.e. Adaptation) – a good sign that it is in the right ballpark – and others were movies I would probably like (Talk to her, 25th Hour).

In fact, I liked it well enough that I recommended it to my wife. What we really need, of course, is a joint recommendation program that accounts for both of our personalities and moods at the same time. Of course, there may not be any light-hearted/political/significant/artsy/funny/deep/martial arts epic/chick flicks out there….

I guess the work-around is for her to sign in and enter her movie mood, and then generate recommendations until I see something I like.