Subscribe to DSC Newsletter

Social recommendation websites potential mob target

Recently Yelp was hit with a class action lawsuit for defamatory and fake bad restaurant reviews. Although the lawsuit was dismissed, the plaintiffs claimed that Yelp would post bad reviews of businesses not purchasing $500/month advertising on Yelp. Much of the discussion was about Yelp's algorithm to automatically identify and filters out reviews that need to be filtered out.

I am wondering whether those posting bad reviews were really associated with Yelp (as the plaintiffs believe), or instead an independent mob organization. Indeed, how could a social recommendation website handle the following scenario (using ad-hoc fraud detection techniques):

  1. A mob decides to launch a new "business product", less risky and less dangerous than traditional mob activities, and even "softer" (less criminal?) than click fraud: extorting money from businesses listed on Yelp or similar websites
  2. They start targeting a few hundreds businesses by posting very bad reviews from various accounts and IP addresses
  3. They contact the victims and offer to clear their profiles for $500/month
  4. The mob operates from abroad, in broad daylight, with a respected website etc.
  5. They might even purchase services from their victims before posting bad reviews, to make their reviews legitimate

For now, I believe this new extortion business model is not easily scalable - less than click fraud at least. So I don't think the mob will quickly integrate it in its business operations, not until they figure out a way to efficiently scale it, possibly using Botnets or trained workers in India or Romania. Yet it could attract a large number of newly graduates as they can't find a job that pays back their student loan, or some highly skilled  laid-off workers who have been unemployed for a long time. For a fresh smart graduate, it could wok like this:

  • Target 500 well selected businesses with the right sales pitch
  • Outsmart fraud detection algorithms
  • Get 50 out of 500 businesses to accept your scheme (to get rid of bad reviews) and charge them $100/month
  • Make $60,000/year out of your scheme.

If only 100 people implement this idea, then about 50,000 businesses (those most likely to pay extortion fees) would be impacted by these bad reviews. 

Views: 349

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Carla R. Ackley on August 13, 2012 at 7:08am

I read of a Reputation company that was doing this and then hitting people up for more money saying they would post negatives on them if they didn't keep paying. I saw an article about it recently, plus there are a lot of bad reviews on reputation companies online when you do a seach in Google for "reputation company scam".

I personally do not like the review sites model as it's inheritantly flawed. I have seen many negatives placed on clients that don't seem to know who these "customers" were, and we suspect they are competitors, or someone paid to post bad reviews, plus there is the chaos factor of just getting an evil person posting a fake review for the heck of it, just to mess with the company that they may or may not even know.

For Yelp, I have seen bad comments posted and then Yelp leaves the bad review up and removes our response to it, and your only option is to sue and most people can't afford the expense. Legal can be a money pit. On Google Places, I've had to remove entire businesses from Google just to get rid of the bad review.

To post a review on Yelp, you just sign up. There is no real validation process. Anyone can create an account.

I think there should be a standard of verification against your drivers license, ss security #, or something else important/legal to track back to. Of course that opens up a whole plethora of other problems. Phone numbers seems to be a popular verification method currently.

Several years ago I made a similar suggestion to eBay about not allowing "free" email accounts for logins as there was too much fraud, and no way to track back to it. I suggested that you had to have a real email address listed. eBay discussed it, but apparently decided not to do anything at that time, and for quite a while. I believe they are doing a telephone # verification now, so it's a step in the right direction, but even that can be hoaxed.

Carla Ackley

Ackley's Custom Site Submission

http://www.classichits.com/

Videos

  • Add Videos
  • View All

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service