Helping to fix the fake news problem with metadata

Obviously there has been a lot of talk recently about fake news, how it may or may not have factored into the recent US elections, what Facebook should or shouldn’t do about it, etc. Even President Obama has weighed in on the issue. So I figured now it’s my turn ;-)

A small but potentially significant thing that we could do is to tag “satirical” content as such using metadata — that is, the “information about the information” which we use to describe things (think of the time, location tags on a photo on your phone, for example).

One growing type of metadata is the schema.org project, sponsored by the major search engines (and using the World Wide Web Consortium’s communities platform to operate) but free for anyone to use. Sites use schema.org metadata to tag content as being restaurant reviews, recipes, events or news stories. Without changing how content looks to human readers, it can be made a whole lot more understandable to computers, which obviously helps with search results, bots answering questions about recipes. And maybe, I thought, it could help people make it clear when articles are intended to be satirical.

So I asked Dan Brickley, Semantic Web guru, Googler and maintainer of schema.org, if there were any plans to make a tag for satirical news part of the schema.org markup. It’s an ever-growing standard and tends to follow industry trends — a recent effort has focused on fact-checking articles which will hopefully provide the right tools to debunk false articles. I’m very happy to say that Dan replied that he had indeed looked into it a few years ago, and that as a result of me asking the question, he has revived his proposal for a SatiricalNewsArticle tag. And it looks like it might gain some traction.

Now you might say that there’s no point in creating a fake news tag because the article’s author must voluntarily state that they are writing satire. It’s true that the tag must be consciously added by the publisher of the article but convincing publishers to use it might not be as difficult as you think — according to the recent Washington Post article on a prominent fake news purveyor, he gives the impression that he wants people to realise, eventually, that they’re reading satire.

More to the point, Google and Facebook have both said that they are banning fake news sites from making money via their advertising networks. A “carrot and stick” approach that bans fake news sites from making money from advertisements, unless the articles are tagged as satire, could be a nice way to manage the problem. I’m not saying that anyone at Facebook and Google are proposing that approach, but if they did, it might work.

To be clear, putting a “SatiricalNewsArticle” tag in your article metadata won’t do anything in itself, but it gives social media sites and search engines the information that they need to include a little “Satire” tag next to the article’s headline if they choose.

If you have thoughts on the SatiricalNewsArticle proposal, please make comments on the schema.org github issue.

It’s a shame that people can’t seem to recognise satire without it being pointed out to them. But that’s the world we live in, so why not try to make it a little better?

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.