If you’re obscuring your identity with a gender-neutral name on Twitter … or flat out lying about it … prepare to be unmasked. The Atlantic Wire’s Rebecca Greenfield reports there’s a new algorithm out there that can figure out, 75.8 percent of the time, whether you’re male or female by what you tweet.
Women evidently tweet a little more expressively. Seriously! I knooooow! Exclamation points, repeated letters, emoticons, expressions of affection and abbreviations make it likely that a woman is doing the tweeting. Such information is useful in marketing … “but,” Greenfield writes, “it could also help identify phonies misrepresenting themselves. Like, say, older men pretending to be lesbian bloggers. Remember when the Gay Girl in Damascus revealed himself as a 40-year-old dude named Tom?
Fast Company’s David Zax reports that the new research comes from the Mitre Corporation, which will be presenting their findings at a conference this week in Scotland and offers an “unintentionally hilarious” chart from the paper showing the most gender-skewed words to turn up Twitter. There are many in the female category, including “hahaha,” “love,” and “etsy,” and two in the male category: “http” and “Google.”
Funnier still is a study by researchers at Johns Hopkins of possessive bigrams, pairings starting with “my” followed by another word, which researchers broke down by gender and, very interestingly, political affiliation. Some of the most-often tweeted possessive bigrams by Democrats included “my” followed by “yoga,” “vegetarianism,” and “upscale.” Republicans tweeted about “my” followed by “zionist,” “weapons” and “walmart.”