Team infiltrates paid 'army' of interest posters

You've studied the blogs and the consumer comments online, and you're ready to hit the mall for that special holiday gift. You’re about to make a major investment based on bullet-proof online research -- right? Think again.

That’s the advice of Cheng Chen, a graduate student in UVic’s department of computer science. Cheng took his master’s research undercover and collected data as a member of China’s “Internet water army,” so-called because these paid posters are hired to “flood” websites with comments and threads concerning products and services.

Cheng discovered a world of cut-throat marketing strategies and a growing job market for high school and college students with few skills, but plenty of time on their hands. With his co-supervisors, UVic computer science professors Drs. Kui Wu and Venkatesh Srinivasan, Cheng designed a systematic approach to identify potential online paid posters.

In a case study concerning a business conflict between two of China’s largest technology companies, Tencent and Qihoo 360, the research team investigated reports and comments on two news websites, Sina.com and Sohu.com.

Cheng’s experience undercover revealed the organizational structure of paid posters and allowed the team to identify patterns of behaviour of normal posters and of paid posters. The team found differences in the kinds of comments posted by members of the “water army” and differences in the percentage of replies, the time of posts, the duration of activity and the number of reports commented on by the posters.

Cheng says “water armies” aren’t limited to China. “Paying posters is popular, not only in China, but in every country,” he says. “The only qualification is a credit card [to establish a user ID.]”

The researchers next plan to extend their investigation to the popular Yahoo! Answers website. Also in their sights is a business venture of their own–the researchers are developing an online service to help consumers do their own detection work.

The team’s report, “Battling the Internet Water Army: Detection of Hidden Paid Posters,” can be read at: arxiv.org/abs/1111.4297