artificial-intelligence

Google harnesses machine learning to help publishers thwart abusive online comments

Google and its early-stage incubator, Jigsaw, have launched a new tool that uses machine learning to help publishers combat online abuse.

With Perspective, Google is offering online publishers an API they can integrate into any of their platforms that facilitate user comments. Perspective taps a human-generated database of comments that have already been tagged as abusive or toxic. Through the API, publishers can connect their own comments with the hundreds of thousands on Google’s database, and Perspective then rates them based on how similar they are to those previously flagged. Perspective facilities corrections from users, too, so it should improve as it receives more feedback from people using it.

The Perspective API essentially provides the data, and publishers can elect to use this data in different ways. Comments could be flagged by publishers to be manually reviewed by moderators or by members of the community. Alternatively, a publisher could decide to illustrate to a commenter in real time that what they’re writing is abusive.

Read more…