A new study describes a technique to identify users who dishonestly post online comments, fake reviews, or tweets across various platforms. The practice, commonly known as "astroturfing," will now be a thing of the past as a new algorithm will analyze multiple writing samples to find how true they are.
Developed by Dr. Kim-Kwang Raymond Choo, professor of cybersecurity and information systems at the University of Texas at San Antonio (UTSA), the program can easily spot people who spread propaganda through multiple accounts. Choo and his team discovered that authors have a challenging time to completely hide their style of writing in their text. The new method uses punctuation, context, and word choice to find out whether a single individual or many people are behind the samples.
Aided by two of his past students, Choo used the technique to study writing samples from the most renowned commenters on many news websites. To their astonishment, they found that many people advocating their opinions on the internet were actually all connected to a handful of single writers with many accounts.
Astroturfing is allowable, but there's still a question about its fairness, said Choo. He added that this practice has been in existence since social media started gaining popularity. The technique is being used by businesses to influence online shoppers or social media users by paying an associate to post fake reviews on various sites about products. It's also widely used to create numerous fake accounts to support opinions, which gives the reader a false idea of a consensus when the fact is that it's just a single person manipulating everything.
Businesses can use it for many reasons like sabotaging their competitors or supporting their services or products through fake identities, said Choo. Even the big guys are known to use astroturfing. For instance, the Tea Party Movement, former Pres. Geroge W. Bush, Hillary Clinton, and even Donald Trump have been accused of using this technique to gather their vote base.