One of the great things about the internet is the way in which communicating with other people becomes so much easier. Also, this interaction can lead to productive debates that help to increase the knowledge and understanding of those participating.
Unfortunately, there are also those people who use the anonymity of the net to cause trouble. The most common of these “trolls” are the ones who place pointlessly negative and abusive comments on websites. While it has faltered a bit in popularity, the Times remains a prominent news site that attracts many visitors. While there are are still many worthwhile comments on their stories, there is also an inordinate amount of junk.
The unenviable task of going through this dross has been the responsibility of a team of employees. Would you want to look at 12,000 comments a day? Probably not, and I suspect that most of these people would rather be in another department.
Regardless, the Times has prided itself on making sure that its comments sections emphasize quality over quantity. That approach continues with their addition of Moderator, a new technology that will aid in the fight against trolls.
Moderator uses machine learning technology to prioritize those comments that are more worthwhile over those that are merely spam or pointless and combative. Moderator is also sophisticated enough that it can automatically approve some comments without the need for human input.
In short, Moderator scans the comment and awards it a summary score based on the likelihood that one of the Times human moderators would reject it. The employees then use that score to prioritize those comments. The paper hopes that the reduction of negativity will help to spur reader engagement, which will also help their traffic and advertising revenue.
The Times’ full statement on their addition of Moderator can be found here.