BRUSSELS – Alphabet’s Google and subsidiary Jigsaw Thursday launched a new technology to help news organisations and online platforms find malicious comments on their websites.
The technology, named Perspective, will review comments and analyse them based on how similar they are to comments people said were “toxic” or likely to make them leave a conversation.
It has been tested on the New York Times and the companies hope to extend it to other news organisations such as The Guardian and The Economist as well as websites.
“News organisations want to encourage engagement and discussion around their content, but find that sorting through millions of comments to find those that are trolling or abusive takes a lot of money, labour, and time. As a result, many sites have shut down comments altogether,” Mr Jared Cohen, President of Jigsaw, which is part of Alphabet, wrote in a blog post. “But they tell us that isn’t the solution they want. We think technology can help.”
The artificial intelligence examined hundreds of thousands of comments that had been labelled as offensive by human reviewers to learn how to identify the potentially abusive words.
CJ Adams, Jigsaw Product Manager, said the company was open to extend the technology to all platforms, including larger ones such as Facebook and Twitter where trolling can be a major issue.
The technology could in the future be expanded to trying to spot personal attacks or off-topic comments too, Mr Cohen said.
Perspective will not decide what to do with comments it finds are potentially abusive; rather publishers will be able to label them to their moderators or develop tools to help commenters understand the impact of what they are posting on websites.
The initiative against trolls follows efforts by Google and Facebook to combat fake news stories in France, Germany and the United States after they came under fire during the US presidential vote when it became clear they had inadvertently fanned false news reports.
The debate surrounding fake news has led to calls from politicians for social networks to be held more liable for the content posted on their platforms.
The Perspective technology is still in its early stages and “far from perfect”, Mr Cohen said, adding he hoped it could be rolled out for languages other than English too.