In Conversation With….

…Del Harvey, the Head of Twitter’s Trust and Safety division.  Yes, really.  I should have included this in my last post, but so much happened in the run-up to the Easter weekend that I decided to leave it for a separate piece.

A colleague sent me a link to an article outlining how Twitter  was setting up a council to look at abuse on its platform, so I sent her a Tweet, she responded, we swopped email addresses, and then she phoned me and we chatted for twenty minutes.  I’d previously sent her a copy of the paper I submitted for the Web Science conference  as it was all about the problems of identifying misogynist abuse on Twitter, which I think she may have skimmed through.  She was certainly aware of the issues I’d raised, as I would expect.

Basically, it’s very difficult to identify an abusive Tweet.  One hundred and forty characters, even if they’re a) written TO an individual and b) contain offensive language don’t necessarily the recipient regards them as offensive or is insulted.  The meaning of words can change depending on the context in which it is used.  A machine can be programmed to identify offensive tweets by looking for particular words, and it can even be told to look for tweets before and after the one that’s been flagged as potentially offensive, but it still can’t be relied on to make the same decisions a human would as to whether further action is needed.  Indeed, even humans don’t always agree.  Another approach would be to ignore meaning and just focus on identifying combinations of words that appear in tweets most people would agree were offensive.  I proposed tweets like this should be immediately quarantined so that they never make it online.  Yes, some people will be cross but given some of the things that are said and threats made to women on Twitter, I for one wouldn’t care.  But anyway.

Del, and Twitter, prefer to look at this problem as one of managing the way users interact with Twitter, suggesting things such as disincentivising people from being abusive, and/or giving users who receive abuse more control over who can see their tweets and engage with them.  Certainly, there’s research to show that the bullying on Twitter – which seems vicious at times – is regarded as nothing more than a game by the perpetrators.  I don’t suppose it feels like that if you’ve had someone suggest you sit on a butcher’s knife, though.  Indeed, if tweets were quarantined on the basis that a particular combination of words was used, in no time a game would develop to try and see how the quarantine algorithm could be outwitted.  It happened on Reddit (certain words were automatically substituted with less offensive ones in posts.  This included things like references to supporting Hitler), it’ll happen on Twitter.  Sadly, social media has exposed a very nasty form of behaviour certain people just can’t resist indulging in, and even more sadly can’t see why what they’re doing is so abhorrent.  Or won’t see.  Tackling that kind of problem is one for society.

We talked over all these things, and more.  It was amazing to have the ear of someone at the top of a thing that touches all our lives (well, those of use who use social media.  Which is you, if you’re reading this).

Finally, I asked her if she had any ‘tips’ for future research students regarding under-explored areas…. Naturally, Del couldn’t give too much away, but she mentioned these things:

  • Connections and interactions, especially ‘mentions’ and ‘follows’.
  • Predicting spikes and shifts of interest across networks.
  • Are there direct relationships between different communities?  Do different communities interact – where are the edges of a network, and what other communities touch it?

Most people use social media like Twitter, or Facebook, or Tumblr, or Instagram, without a moment’s thought of the wider implications of what they’re doing.  Researchers like me, though, are fascinated with what people do online, and how they’ve adapted their behaviour to the new medium.  Some  of it is negative, like misogynist abuse on Twitter, but much more of it is positive, such as crowdfunding or online protests.  I’m interested in what teachers say on Twitter – there appears to be an extensive and very active network of people expressing their opinions about all things educational.  This network has developed in a way it never could have without the web and a social media platform like Twitter, and yet unless you’re a part of it, you might never realise it’s there.  Bringing it to the fore of academic research will be my life for the next two-and-a-half years.

Thank you, Twitter, for remaining open enough for researchers like me to learn something about what humans do on the web; and thank you Del for sparing some of your precious time to talk to me.

p.s. the paper has been accepted as a extended abstract and a poster.  More on that later…


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s