The team of computer scientists looked for patterns that might indicate paid posting, though Wu was careful to say they hadn't found a way to definitely prove which comments were paid and which were not. With a sample size of over 500 users and 20,000 comments from the news sites Sina.com as well as 200 users and over 1,000 comments from Sohu.com, their findings revealed a markedly high rate of paid comments on these sites, but since they weren't able to track comments posted from smartphones — evidently all mobile posts show up under the same username — they weren't able to determine exactly how many of the comments came from paid posters. But it's a lot. "Frankly we have no idea how many people are in the army; it's probably very big," Wu told The Atlantic Wire. "This is a kind of underground business."
Wu went on to explain that they suspected the practice was also widespread. Pointing to The Guardian's coverage of a United States military operations to target social media with fake online personas, he suspects that North America was a likely outpost for the Internet water army, setting up the suggestion that the U.S. Army might even be supporting the practice. They plan to expand the scope of the research but for not it's still in the early stages. They haven't been able to trace the activity to specific users very well — they do know that there are clandestine companies that support the practice — but the behavior patterns of the paid posters are fascinating and might help cyber security firms build better spam-catching software. MIT's Technology Review explains:
They discovered that paid posters tend to post more new comments than replies to other comments. They also post more often with 50 per cent of them posting every 2.5 minutes on average. They also move on from a discussion more quickly than legitimate users, discarding their IDs and never using them again.
What's more, the content they post is measurably different. These workers are paid by the volume and so often take shortcuts, cutting and pasting the same content many times. This would normally invalidate their posts but only if it is spotted by the quality control team.
So just in case you weren't worried enough about the drone takeover, Wu says it will take some time to figure out how to tell the difference between genuine users and paid spam soldiers. "It's very hard to prove unless the user admits he is a paid poster," Wu said.
Unfortunately, it gets worse. As we learned from Gawker's Hamilton Nolan and the case of the cash for links practices of some marketing agencies, this kind of paid information twisting is becoming increasingly common not just amongst commenters but bloggers themselves, and there's nothing we can do about it — unless you're a blogger, in which case you should just use common sense. In some ways, paid content is inevitable. To use Nolan's phrase: "Advertisers want to advertise. They don't particularly care about ethical journalism; it's not their job."
This article is from the archive of our partner The Wire.