Google is getting serious about making the internet a safer, friendlier place. Their new tech, called Perspective, takes aim at nothing less than the nastiest of online nasties: the comments section.
Recognising that trolls are a mainstream problem
The comments section of websites, in particular those of online news outlets, often showcase the very worst of the internet’s ability to facilitate abuse and harassment.
Since the 2016 US Presidential election, trolling behaviour and general online ugliness have been attracting a serious mainstream coverage.
The idea that the web-based media could help to drive meaningful conversation and enlightening debate is quickly being replaced by the sinking realisation that people don’t seem to be up to the task of maintaining civility when granted a mask of anonymity.
The problem is so bad that some major news providers have abandoned their comments sections altogether.
In what seems to be a direct response to public concerns, Google’s internet safety offshoot Jigsaw has launched a new API called Perspective—a sophisticated bit of code that developers can use to better moderate their comments sections.
Perspective uses machine learning to grade comments for how toxic they are to the overall discussion. It goes beyond simply flagging offensive terms and abusive language to delve into the conversational context.
By facilitating more nuanced moderation—particularly for websites with major traffic where doing so without automated assistance may be impossible—the hope is that publishers can use the API to help them maintain better conversations.
To learn more about how Perspective works and how developers can put it to use, click here.