In China, they’re called the Internet Water Army: legions of people paid to flood online hangouts with postings and comments (primarily for marketing purposes). And according to academic researchers, they’re degrading information quality on the Web.
In a research paper released this week, researchers Cheng Chen, Kui Wu, Venkatesh, Srinivasan, and Xudong Zhang explain that paid posters are hired by public relations and marketing firms to post content on Websites–usually about a social event, product, or company.
For marketers, paid posters are a way to control word-of-mouth publicity about whatever they’re trying to sell.
“If a company hires enough online users, it would be able to create hot and trending topics designed to gain popularity,” the researchers write. “Furthermore, the articles or comments from a group of paid posters are also likely to capture the attention of common users and influence their decision.”
That which is good for marketers, however, isn’t always good for consumers.
“Though and interesting strategy in business marketing, paid posters may create a significant negative effect on the online communities, since the information from paid posters is usually not trustworthy,” the paper states.
The practice can poison the attitudes of web surfers, especially when the two armies clash in cyberspace. “When two competitive companies hire paid posters to post fake news or negative comments about each other, normal online users may feel overwhelmed and find it difficult to put any trust in the information they acquire from the Internet.”
To illustrate this, the researchers cited an instance in which a seemingly inane message was posted to a World of Warcraft forum in China.
“Junpeng Jia, your mother asked you to go back home for dinner!” The message said.
In two days, the message had garnered over 300,000 replies and seven million clicks. It was later revealed that a PR firm was behind the surge in traffic–it wanted to maintain interest in the site while it was down for system maintenance, so it hired 800 posters who used 20,000 identities to create the illusion of purposeful activity happening there.
Although the researchers focused their attention on paid posters in China, these techniques are used all over the world. The U.S. military, for example, has been working on software to automate the process of creating “sockpuppet” armies to invade social networks and online forums and gather info on terrorists and terrorist organizations. The researchers also showed how “socialbots” made up of bogus “friends” could be used for mischief on Facebook.
Through their research, the scientists identified several ways to identify potential paid posters on a forum. Paid posters add new comments to a forum more often than typical members, for example. This is because they work under tight time constraints and don’t have time to read other member’s comments.
For the same reason, the rate between comment postings is shorter, too. And they don’t use an I.D. very long, either. Their “mission” time is usually short. When their mission is done, they discard their I.D. and never use it again.
Detecting infiltration of a website by paid posters with automated systems that use semantic analysis can be effective, the researchers noted.
“The reason why the semantic analysis improves performance is that online paid posters often try to post many comments with some minor edits on each post, leading to similar sentences,” they explained. “This helps the paid posters post many comments and complete their assignments quickly, but also helps our classifier to detect them.”