Back at school I was fond of a play the title of which I cannot even remember now where a character cheekily asks another: “Can you tell if I am black from my handwriting?”. There was an egalitarian insouciance built into that line which appealed to the logic-geek in me and, since then I have extended and modified that line to apply to my online world. I have always wondered whether someone could tell whether I was male or female just from my writing.
Hemingway might have subscribed to the ‘masculine’ school of thought when it comes to penning a line but I have always considered that when it comes to what we write and, even more wondrously, to the stripped-down-to-the-bare-minimum, what we Tweet, genders are invisible.
Well, apparently not, it seems. Researchers from the Mitre Corporation have apparently cracked the gender code to online writing and found a way which allows them to spot the gender of a person, with a high degree of accuracy, from just a single Tweet.
“By scanning for patterns in all the tweets of a given user, Mitre's program was able to guess the correct gender 75.8% of the time--a 20% improvement over the baseline. And even just by analyzing a single tweet of a user, it was right 65.9% of the time--an over 10% improvement over the baseline.”
When more than one datafield was sampled the accuracy of the results rose to a whopping 92%.
Apart from the fact that this makes it harder to hide anonymously behind chat and pretend to be a Californian babe when in truth you’re a hirsute, middle-aged gentleman, typing from Islamabad, this has important ramifications from marketers who continuously seek to find new ways to gauge the demographic make up of their audience and reach of their Tweets.