Turning to the machines: New algorithm uses ML to identify cyber-bullies on Twitter
The algorithms 'learn' how to tell the difference between bullies and typical users by weighing certain features.
For the study published in the journal Transactions on the Web, the research team analysed the behavioural patterns exhibited by abusive Twitter users and their differences from other users.
"We built crawlers - programs that collect data from Twitter via variety of mechanisms," said study researcher Jeremy Blackburn from Binghamton University in the US.
"We gathered tweets of Twitter users, their profiles, as well as (social) network-related things, like who they follow and who follows them," Blackburn said.
The researchers then performed natural language processing and sentiment analysis on the tweets themselves, as well as a variety of social network analyses on the connections between users.
Facebook, Twitter Instagram: Tips & Tricks To Keep Social Media Private
They developed algorithms to automatically classify two specific types of offensive online behaviour, i.e. cyber-bullying and cyber-aggression.
The algorithms were able to identify abusive users - who engage in harassing behaviour like those who send death threats or make racist remarks - on Twitter with 90 per cent accuracy.
"In a nutshell, the algorithms 'learn' how to tell the difference between bullies and typical users by weighing certain features as they are shown more examples," said Blackburn.
"Our research indicates that machine learning can be used to automatically detect users that are cyber-bullies, and thus could help Twitter and other social media platforms remove problematic users," Blackburn added.