Fuck, fake reviews will get much harder to get away with...

LMSInc.

LMSInc.

New member
Oct 21, 2009
325
5
0
Software picks out fake online reviews

Based on a subset of the 800 reviews, the team created a fake-review-detecting algorithm. When used in a way that combined the analysis of keywords and word combinations, that algorithm was able to identify deceptive reviews in the entire database with 89.8 percent accuracy.
Click to expand...

Looks like it's still in its infancy, but I guess it's only a matter of time until it is rolled out into larger sections of the web... Wonder if G will adopt something like this for Places?

That would make my business a little harder.
 


Looks like the algorithm is so far only able to identify dumb reviews (or reviewers) not fake ones. Maybe in a hundred years...
 
I personally want to see this thing when it's done and fuck around with it to see what it can and can't do.

I guess if you really think about it, at the core of how it works, it couldn't be too far off from something like Copyscape, so I suppose once something like this is released, it is only a short amount of time before we can figure out what triggers flags, and adjust accordingly.

When I think about it in that light, maybe this could be a good thing for those of us that fiddle with reviews quite often.

It will seed out the shitty, broken-English reviews and thus leave just the legit ones, and our top-notch ones as well. :)
 
This test was done on a very small scale for a very specific niche, I wouldn't get too worried yet.

Although I'm sure for many niches it would be fairly easy to make since most people are not creative and just copy the same wording everyone else uses.
 
medway said:
This test was done on a very small scale for a very specific niche, I wouldn't get too worried yet.

Although I'm sure for many niches it would be fairly easy to make since most people are not creative and just copy the same wording everyone else uses.
Click to expand...

Precisely why I actually like the idea of this.

It may very well have a similar effect to that of Craigslist once they implemented all of their account/spam measures heavily. I was involved with CL spamming at the time, and it pushed out so much of the competition it was unreal. It made shit so much easier for those whom were dedicated.
 
It's still a piece of shit, if you really look at the comments about the guy that responded they assumed every review from hotels.com was legit and their "good" reviews were TripAdvisor. That's like saying you can evaluate the knowledge of an internet marketer by their DP or WaFo account alone.
 
After the human trials, the researchers then applied statistical machine learning algorithms to the reviews, to see what was unique to both the genuine and fraudulent examples. It turns out that the fake ones used a lot of scene-setting language, such as "vacation," "business" or "my husband." The genuine ones, on the other hand, tended to focus more on specific words relating to the hotel, such as "bathroom," "check-in" and "price."
Click to expand...

Yeah that is going to be really impossible to fool.
 
Dumbest shit ever. Worst case scenario is that everyone gets good at writing fake reviews.
 
wow i agree ppl will start adapting and write better fake reviews but you never know these days with technology look at Panda smashing apart SERPS and honeypot over taking forum spam....nothing beats manual labor work hard for that review
 
You must log in or register to reply here.
Top Bottom