Collaborative filtering would be cool, yes. However the situation for netflix is a bit different than restaurants. It's hard to get really good overlap with a large number of people for a lunch that exists only in one specific location, compared to a film which can be viewed anywhere in the world, at any time, by anyone (and then encouraged to review by their direct point of sale (Netflix)). Also the cost (mostly time) in generating a lot of reviews by a single person to establish any kind of baseline is much higher.
And, on top of that, it doesn't help the random anonymous user with 0 reviews submitted as much.
Someone should get on that though, because if you could solve the problem of how to set up proper recommendation systems for food, that'd be awesome.
Who knows, maybe Yelp as experimented with this? I don't know what kind of logged-in userbase they have though. Maybe if you had something integrated with the restaurants, where people could be logged in via their smartphones, and give a review to some NFC device at the restaurant. Lower the barrier of reviewing, encourage reviews right at the point of sale (similar to reviewing on netflix.com).
Someone must have already thought of this before though.
And, on top of that, it doesn't help the random anonymous user with 0 reviews submitted as much.
Someone should get on that though, because if you could solve the problem of how to set up proper recommendation systems for food, that'd be awesome.
Who knows, maybe Yelp as experimented with this? I don't know what kind of logged-in userbase they have though. Maybe if you had something integrated with the restaurants, where people could be logged in via their smartphones, and give a review to some NFC device at the restaurant. Lower the barrier of reviewing, encourage reviews right at the point of sale (similar to reviewing on netflix.com).
Someone must have already thought of this before though.