These days, a lot of the guesswork has been taken out of travel, especially when it comes to choosing hotels. Sites such as Hotels.com and TripAdvisor.com have very active communities of customers who reveal issues that never show up on websites or in advertising. You can get a clue about how unresponsive or rude the staff is, how filthy the rooms were, how thin the walls were, and how loud that nightclub right next door was -- hotels are now kept on their toes by social media.
Unless, that is, review sites are sprinkled with fake reviews with how wonderful a hotel experience was. Or, for that matter, reviews of any other products or services. Many of these glowing reviews seem credible -- can they be spotted?
Yes, they can be spotted, say researchers from Cornell University. The team, led by Myle Ott, developed an algorithm which they claim can spot deceptive text with 90% accuracy. On average, it's a 50-50 proposition whether people can spot such deception.
CNET's Eric Smalley provides an analysis of the deception-spotting tools, noting that the Cornell system is similar to software that sniffs out plagiarism. The algorithm searches for certain clues, just as people telling lies inadvertently give off clues such as not looking their listeners in the eye, or giving off telltale twitches:
"In part, deceptive writers used more verbs than real review writers did, while the real writers used more punctuation than the deceptive writers. The deceptive writers also focused more on family and activities while the real writers focused more on the hotels themselves."
The ability to weed out planted opinions assures the credibility of the social media review system.
This post was originally published on Smartplanet.com