4 ms·

How can a computer evaluate the 'realness' of a review? The idea of Yelp, crowdsourced business reviews, is great, but do we really prefer trusting super active

How can a computer evaluate the 'realness' of a review? The idea of Yelp, crowdsourced business reviews, is great, but do we really prefer trusting super active web content contributors, seeking cyberspace attention, over experts (eg Zagats)?

ddw 15y ago
I think they do things like filter out reviews from a user that has only posted one review, or if your name appears to be fake, etc. But who knows how exactly their algorithm works. They're afraid of competitors leaving negative reviews for a business and there's probably something to that, but there's no transparency about it. And they've been doing the extortion thing for a while now: http://dealbook.nytimes.com/2010/03/19/yelp-under-fire-from-lawsuits-citing-ad-pressure/?ref=yelp http://dealbook.nytimes.com/2010/03/19/yelp-under-fire-from-... I always thought it would be great if there was a Yelp for landlords - because landlords are rarely held accountable and it's not like you ask for references before you sign a lease - but there would probably be too many false reviews posted to make it work or tenants wouldn't feel comfortable speaking freely (because unfortunately tenants still need references). There's actually a website trying it, but doesn't seem like it's going anywhere right now: http://www.donotrent.com http://www.donotrent.com Hopefully people will see Yelp for what it is: a small sample of the population that may be dissimilar to you writing probably biased reviews.
- ddw 15y ago
  Why downvoted?
patio11 15y ago
From the perspective of a web searcher, a SERP with reviews for four [Buffalo Grove dentists] written by amateurs beats a SERP with zero reviews written by experts. From the perspective of a large advertising firm, consistently replicating that at scale gives you a lot more inventory to sell. One could say something similar regarding content farms and big daddy G.
pbreit 15y ago
Zagats is primarily crowdsourced.

ddw 15y ago

I think they do things like filter out reviews from a user that has only posted one review, or if your name appears to be fake, etc. But who knows how exactly their algorithm works. They're afraid of competitors leaving negative reviews for a business and there's probably something to that, but there's no transparency about it. And they've been doing the extortion thing for a while now: http://dealbook.nytimes.com/2010/03/19/yelp-under-fire-from-lawsuits-citing-ad-pressure/?ref=yelp http://dealbook.nytimes.com/2010/03/19/yelp-under-fire-from-... I always thought it would be great if there was a Yelp for landlords - because landlords are rarely held accountable and it's not like you ask for references before you sign a lease - but there would probably be too many false reviews posted to make it work or tenants wouldn't feel comfortable speaking freely (because unfortunately tenants still need references). There's actually a website trying it, but doesn't seem like it's going anywhere right now: http://www.donotrent.com http://www.donotrent.com Hopefully people will see Yelp for what it is: a small sample of the population that may be dissimilar to you writing probably biased reviews.

ddw 15y ago
Why downvoted?

patio11 15y ago

From the perspective of a web searcher, a SERP with reviews for four [Buffalo Grove dentists] written by amateurs beats a SERP with zero reviews written by experts. From the perspective of a large advertising firm, consistently replicating that at scale gives you a lot more inventory to sell. One could say something similar regarding content farms and big daddy G.

pbreit 15y ago

Zagats is primarily crowdsourced.