TechSpot

Report shows 50% of misogynistic tweets come from women

By midian182
May 26, 2016
Post New Reply
  1. For as long as there has been an internet, there has been online abuse. While this covers everything from racism, anti-Semitism, and general cyberbullying, one area that has come under the spotlight in recent years is sexism, especially in the wake of Gamergate.

    In the UK, five MPs - Yvette Cooper, Maria Miller, Stella Creasy, Jo Swinson and Jess Phillips - have launched the Reclaim the Internet campaign, which wants to examine ways of making the internet less aggressive, racist, homophobic, and sexist.

    As part of the launch, the campaign has released data from a three-week Twitter study by think tank Demos. It found that there were more than 200,000 aggressive tweets that used the words “****” and “*****” sent out to 80,000 people. While it’s a sad fact that this level of abuse isn’t too surprising these days, what does come as a shock is that over half the tweets were sent by women.

    It’s important to note that Demos used special natural language-filtering algorithms that were able to identify when these terms were directed at someone in an explicitly aggressive, derogatory fashion. It avoided instances of "self-identification, and those that were more conversational in tone or commenting on issues related to misogyny (ie. referring to ‘**** shaming’, ‘**** walks’)"

    As discovered by The Guardian’s ‘the web we want’ project, articles penned by women consistently receive more abusive comments than those written by men. It was assumed that most of these remarks came from males, but Demos' research suggests that misogynistic abuse often originates from both genders.

    Speaking about Demos’ study, researcher Alex Krasodomski-Jones said it was not about "policing the internet" but was more "a stark reminder that we are frequently not as good citizens online as we are offline."

    Image credit: Soze Soze / Shutterstock.com

    Permalink to story.

     
  2. Take that sarkeesian.

    But obviously the reports were conducted by male misogynistic scumbags right?
     
    Hexic, SirChocula and wastedkill like this.
  3. Hyrax

    Hyrax TS Enthusiast Posts: 17   +7

    Let's try to keep the discussion serious and not bring in people like sarkeesian. These type of people destroy any chance at a constructive and grounded discussion.
     
    mbrowne5061 likes this.
  4. SirChocula

    SirChocula TS Booster Posts: 75   +59

    That's not very logical. Sarkleesian and her likes are the people who are up in arms about misogyny, cultural appropriation, SJW lunacy, etc. It is very serious to them and she has even spoke before the UN over the issues of internet censorship, etc. If ordinary and sane feminists were are the forefront, we would have a civil debate and the country wouldn't be in the social f*ckfest/upheaval that it is currently.
     
  5. chollatta

    chollatta TS Rookie

    If 50% are women then the title could of just of easily been 50% come from men Or Misogynistic tweets are equal between male and females.
     
  6. davislane1

    davislane1 TS Evangelist Posts: 3,558   +2,362

    200,000 aggressive tweets that used the words “****” and “*****” sent out to 80,000 people. While it’s a sad fact that this level of abuse isn’t too surprising these days, what does come as a shock is that over half the tweets were sent by women.

    How is this even remotely curious, let alone shocking? '$|ut' and 'wh0re' have been basic labels for centuries, especially when used by women. In fact, women are far more likely to use those terms IRL than men simply because the fact that HR won't crucify them for it.

    This whole story falls under the category of "grow a spine." If you can't handle mean comments on the internet, you're too short for the ride.
     
    p51d007 likes this.
  7. Hyrax

    Hyrax TS Enthusiast Posts: 17   +7

    Emma Watson has been quite sane and ordinary on this topic if I remember correctly.

    The point was exactly that if we want to move forward we have to stop giving floor or any attention to people who use dishonest ways and even lie to support their views. So any mention of these people should be avoided or people will start to think they have some kind of insight or credibility in the matter
     
    mbrowne5061 likes this.
  8. Camikazi

    Camikazi TS Maniac Posts: 817   +231

    Did we need a report for this? I've known for years that women are the people who are the hardest and meanest towards other women, they are vicious honestly.
     
    Agnomen and davislane1 like this.
  9. davislane1

    davislane1 TS Evangelist Posts: 3,558   +2,362

    There is nothing more harmful to women as a whole than other women.
     
  10. cliffordcooley

    cliffordcooley TS Guardian Fighter Posts: 8,556   +2,900

    Nothing new here move along. Seriously did anyone think differently?
     
  11. AsiAsi

    AsiAsi TS Rookie


    Yes but all the 3rd wave femtards and lying Liberal media say 99% of it is Men, that is why 50% of it being women is relevant
     
    davislane1 likes this.
  12. captaincranky

    captaincranky TechSpot Addict Posts: 11,702   +1,886

    This just in, women comprise 50% (or thereabouts) of the population. The law of averages therefore dictates that I would agree with, and enjoy, 50% of their tweets others might call misogynistic! :mad:

    By the way, WTF does "MP" stand for? Major Pain" in the a**? Who do those b****es think they are? They need to tweet the **** off!
     
    Last edited: May 26, 2016
  13. moftenwrong

    moftenwrong TS Rookie

    For those who want to read the original report by Demos, rather than a summary, see here: http://www.demos.co.uk/files/MISOGYNY_ON_TWITTER.pdf?1399567516.

    This report and TechSpot's summary have some issues:
    - First, the Demos report relies heavily upon self-reported "male" and "female" and has no ground-truth. People can pretend to be whomever they want to be online, 'nuff said.
    - Second, the report's NLP algorithm for detecting a tweet's class as abusive, playful, etc. relies on only 200 labeled messages. This is far too small. There are no cross-validation or clear testing procedures to ensure the data classes are reliable. The quality of some figures like Figures 3 and 7 further undermines the authors' credibility.
    - Detecting sentiment from tweets is tough due to their short length. The report doesn't even both applying, citing, or reviewing more mature sentiment detection algorithms, for example https://www.csc.ncsu.edu/faculty/healey/tweet_viz/tweet_app/
    - The study states "Women are as almost as likely as men to use the terms ‘****’ and ‘*****’ on Twitter." TechSpot translates this to "what does come as a shock is that over half the tweets were sent by women." An interesting reporter's jump from <50% to >50%.
    - A more careful statement might be that 'over half the tweets were sent by people who said they were women, and the majority of them were commenting on news articles or using the terms conversationally, though the Demos report didn't give clear a breakdown of these categories by gender. It's also unclear from the Demos report whether men or women were more likely to use the terms 'rape' or '*****' abusively.'
     

Similar Topics

Add New Comment

You need to be a member to leave a comment. Join thousands of tech enthusiasts and participate.
TechSpot Account You may also...