updated 12:40 pm EDT, Wed April 25, 2012
Google funds user review algorithm research
A new algorithm, partly funded by Google, is aimed at spotting spam in user-generated online reviews. Shown at the World Wide Web 2012 conference in France, GSRank analyzes groups of reviewers and language similarities to identify illegitimate reviews of goods and services.
The algorithm monitors a number of factors to identify the work of a paid reviewing team, including tendencies for the group to post reviews in quick succession on the same product early in its product life, deviating from the general opinion of potentially legitimate reviews, similar content in the reviews themselves, and the amount of content that the group appears to have worked on together.
Consumer reviews are taken into account more often online, highly influencing customer choices. Algorithms such as GSRank are designed to help curb the spamming trade that can easily damage a brand in an extremely valuable marketplace, such as Amazon. Google collaborated with the University of Illinois at Chicago and partly funded the research through a Google Faculty Award. [via ZDNet]