• Subscribe

Taking threats out of Twitter

Speed read
  • Twitter is rife with threats against women but the burden of reporting lies with the abused
  • New machine-learning model identifies misogynistic context with 75% accuracy
  • Model could be deployed at platform level to remove abusive content

Online abuse targeting women, including threats of harm or sexual violence, has proliferated across all social media platforms. But now researchers at Queensland University of Technology (QUT) in Australia have developed a statistical model to help drum it out of the Twittersphere.

<strong>Abuse directed at women</strong> is common on social media. A new algorithm could identify such language and remove it before it reaches and harms the user. Associate Professor Richi Nayak, Professor Nicolas Suzor, and research fellow Dr. Md Abul Bashar from QUT have developed a sophisticated and accurate algorithm to detect these posts on Twitter, cutting through the raucous rabble of millions of tweets to identify misogynistic content.

The team, a collaboration between QUT’s faculties of Science and Engineering and Law and the Digital Media Research Centre, mined a dataset of 1 million tweets then refined these by searching for those containing one of three abusive keywords - whore, slut, and rape.

“At the moment, the onus is on the user to report abuse they receive. We hope our machine-learning solution can be adopted by social media platforms to automatically identify and report this content to protect women and other user groups online,” said Professor Nayak.

The key challenge in misogynistic tweet detection is understanding the context of a tweet. The complex and noisy nature of tweets makes it difficult.

“On top of that, teaching a machine to understand natural language is one of the more complicated ends of data science: language changes and evolves constantly, and much of meaning depends on context and tone.

“So, we developed a text mining system where the algorithm learns the language as it goes, first by developing a base-level understanding then augmenting that knowledge with both tweet-specific and abusive language.

“We implemented a deep learning algorithm called Long Short-Term Memory with Transfer Learning, which means that the machine could look back at its previous understanding of terminology and change the model as it goes, learning and developing its contextual and semantic understanding over time.”

<strong>Protecting women.</strong> Associate Professor Richi Nayak of QUT’s School of Computer Science and Centre for Data Science developed a text-mining system that can differentiate between abuse, sarcasm, and friendly use of aggresive language. Courtesy QUT. While the system started with a base dictionary and built its vocabulary from there, context and intent had to be carefully monitored by the research team to ensure that the algorithm could differentiate between abuse, sarcasm, and friendly use of aggressive terminology.

“Take the phrase ‘get back to the kitchen’ as an example—devoid of context of structural inequality, a machine’s literal interpretation could miss the misogynistic meaning,” said Professor Nayak.

“But seen with the understanding of what constitutes abusive or misogynistic language, it can be identified as a misogynistic tweet.

“Or take a tweet like ‘STFU BITCH! DON'T YOU DARE INSULT KEEMSTAR OR I'LL KILL YOU’. Distinguishing this, without context, from a misogynistic and abusive threat is incredibly difficult for a machine to do.

“Teaching a machine to differentiate context, without the help of tone and through text alone, was key to this project’s success, and we were very happy when our algorithm identified ‘go back to the kitchen’ as misogynistic—it demonstrated that the context learning works.”

The research team’s model identifies misogynistic content with 75% accuracy, outperforming other methods that investigate similar aspects of social media language.

“Other methods based on word distribution or occurrence patterns identify abusive or misogynistic terminology, but the presence of a word by itself doesn’t necessarily correlate with intent,” said Professor Nayak.

“Once we had refined the 1 million tweets to 5000, those tweets were then categorized as misogynistic or not based on context and intent, and were input to the machine learning classifier, which used these labelled samples to begin to build its classification model.

Sadly, there’s no shortage of misogynistic data out there to work with, but labelling the data was quite labor-intensive.

Professor Nayak and the team hoped the research could translate into platform-level policy that would see Twitter, for example, remove any tweets identified by the algorithm as misogynistic.

“This modelling could also be expanded upon and used in other contexts in the future, such as identifying racism, homophobia, or abuse toward people with disabilities,” she said.

“Our end goal is to take the model to social media platforms and trial it in place. If we can make identifying and removing this content easier, that can help create a safer online space for all users.”

Read more:

Read the original article on QUT's website.

Join the conversation

Do you have story ideas or something to contribute? Let us know!

Copyright © 2023 Science Node ™  |  Privacy Notice  |  Sitemap

Disclaimer: While Science Node ™ does its best to provide complete and up-to-date information, it does not warrant that the information is error-free and disclaims all liability with respect to results from the use of the information.


We encourage you to republish this article online and in print, it’s free under our creative commons attribution license, but please follow some simple guidelines:
  1. You have to credit our authors.
  2. You have to credit ScienceNode.org — where possible include our logo with a link back to the original article.
  3. You can simply run the first few lines of the article and then add: “Read the full article on ScienceNode.org” containing a link back to the original article.
  4. The easiest way to get the article on your site is to embed the code below.