Google has released a new API for publishers which uses machine learning to filter out hurtful comments online. The new tool, called Perspective, can be integrated into any publishing platform which utilizes online comments.
Perspective cross-references comments against a human-generated database of comments that have already been labeled as offensive. When a new comment is posted, Perspective will rate it based on how closely it resembles the hundreds of thousands of abusive comments in Google’s database.
If it has been determined that a comment is abusive, Perspective will notify the commenter in real time, which you can see in the example below:
Comment sections have become so volatile as of late that many publishers have opted to remove comment section altogether, rather than trying to moderate them. Perspective aims to make it easier to moderate abusive comments before they go live.
Google revealed Perspective is currently being tested with the New York Times’ comment section. As more sites integrate the API with their publishing platform, Perspective will continue to learn and grow its database of toxic comments, making it even more efficient over time.
Currently, Perspective is only able to flag abusive comments left in English, but that could change over the next year if it ends up catching on with publishers.