Why doesn't Yelp use collaborative filtering?

[mattbell]

Having just moved into the Mission in San Francisco, I had an empty refrigerator but a hankering for food late at night.  "Why not check out the best of the local taquerias?", I thought.  The clear winner on Yelp was this place, with an average of 4.5 stars out of five with a mind-blowing 1100 reviews.  I went there.  The wait was horrendously long, and both the burrito and the soft taco I got (both recommended by reviewers) were horrible.  The meat was far too dry and the burrito tasted soapy.  It could have been an "off" day, but the place was crowded and others seemed to be eagerly eating their meals 

Something can get five stars (or one star) for a lot of reasons.  I'm guessing the 5-star reviewers like the place I tried for being open after the bars close and ridiculously cheap. 

This is where collaborative filtering comes in.  Collaborative filtering is a data analysis technique that analyzes everyone's reviews to find people similar to you.  If I'm picky about food quality but not about ambience, it will find other people who are picky about food quality but not ambience.  Then, these people's reviews can be weighted more heavily to provide more accurate ratings for me.  Netflix uses this technique to provide personalized recommendations of movies. 

I believe that implementing this technique would dramatically increase Yelp's usefulness to its users. 

One nice thing for Yelp about collaborative filtering is that it encourages people to make a lot more reviews.  If the system tells you that you will start getting personalized recommendations once you've reviewed 100 places, you're much more likely to start reviewing.  Right now lots of people are freeloading off of Yelp by reading others' reviews but not posting their own.  Adding collaborative filtering would dramatically change the incentive structure against freeloaders and get Yelp a pile of high quality user-generated content for free.

Another nice thing about collaborative filtering is that it would dramatically weaken the ability of restauranteurs to game the system by posting positive reviews on their own restaurants and negative reviews on their competitors'.  The sham accounts used to make these reviews would carry a lot less weight because the accounts would likely not have enough reviews to influence collaboratively filtered ratings.

Comments

[(Anonymous)]

You should write a proposal to Yelp.

[maradydd.livejournal.com]

I agree. I would help write one.

[mixophrygian.livejournal.com]

Sounds like a great idea- how do you think it could be implemented? Would Yelp have some kind of word-finder for things like "ambience" or "taste" or would the user be able to manually select their priorities from a list separate from the review itself?

[nasu-dengaku.livejournal.com]

Having people manually select their priorities certainly would improve the algorithm, but collaborative filtering could be done totally automatically just with the existing data that people have already entered. There are lots of different ways of doing collaborative filtering, but most involve some automated clustering system that analyzes the similarity between pairs of users.

[bennj.livejournal.com]

If you're willing to pay the extra couple of bucks to get a better burrito, I highly recommend La Taqueria or Papalote. My cheap burrito of choice is at El Toyanese on 24th and Shotwell, but that one is a function of proximity.

[nasu-dengaku.livejournal.com]

I'll *totally* pay a couple of extra bucks for a better burrito. I believe in investing in myself, and the burrito becomes a part of me when I eat it.

Speaking of food, we should meet for dinner again.

if you like investing in yourself, you might also like...

[radium.livejournal.com]

http://www.gracias-madre.com

[icka.livejournal.com]

La Taqueria is where it's at.

[nasu-dengaku.livejournal.com]

Oh, this is way better than Yelp. Thanks!

[radven.livejournal.com]

Taqueria Cancun is the best in SF. I dream of their tortillas...

[serolynne.livejournal.com]

Love your idea... I've had similar issues with Yelp as well.

And seconding the Taqueria Cancun suggestion.

[hansandersen.livejournal.com]

Angyl says that Yelp *already* does this. You need to engage with the Yelp UI and review sites and friend/like other reviewers whose reviews you find helpful, and then it starts weighting the reviews and guidance it gives you.

[nasu-dengaku.livejournal.com]

Hmm... I'm kind of surprised they don't advertise this. Do they just take your friends' reviews and the people whose reviews you've liked into account, or are they also weighting in strangers with similar tastes?

Ironically, my friends often but aren't necessarily the best people to recommend food to me. I can think of one friend in particular who has very different taste in food from me.