El País ombudsman battles the trolls
If you have published online, you have probably dealt with a troll.
People who post off-topic, inflammatory messages distract from constructive debate and have long been a challenge for achieving respectful reader participation in online news websites.
In the most recent one, she noted how difficult is it for the site to filter offensive comments on renowned journalist Iñaki Gabilondo's political videoblog. Pérez Oliva proposed allocating more resources to counter trolls and called for "clear criteria" for comments to halt the nastiness, which include insults about Gabilondo's mother and veer from the oddly violent (discussion of torturing lizards) to the everyday slur "Iñaki sucks."
"I share the view that the newspaper should open as possible to participation...but criticism is one thing and insults and defamation are another," Perez Oliva wrote. She noted that Gabilondo has considered quitting because of the trolls.
"We have taken action, but it takes time to convince readers," said Gumersindo Lafuente, head of digital development at the newspaper. "Many readers still complain about the lamentable tone of some comments. I fear that the speed and scale of the challenge far exceed the capacity" of the current solution.
El País has tried to filter the comments using Eskup, which requires prior registration and bans commenters who do not respect the rules. Still, control is "poor," Perez Oliva said.
In an article titled "The Invasion of the Trolls," Cinthya Sánchez at Mexican daily El Universal says keeping them out is hard because they "change words, separate letters, add numbers and get to say what they want, circumventing security."
Media organizations around the globe are dealing with the troll factor by upping moderation or using tech to filter comments. Some, like The New York Times, have a small staff of part-time moderators.
Others, like The Huffington Post, rely on a 30-person staff of moderators and a computerized system that red-flags words and phrases to create some order out of the chaos of the 4 million comments it receives every month. The semantic analysis engine (also known as JuLiA) was the first acquisition The Huffington Post ever made.