Think tank launches hate speech tracker, targets M’sian social media
The Centre Malaysia has launched a tracker designed to monitor hate speech on Malaysian online spaces.
Operating on machine-learning software, the tracker, called Tracker Benci, is a prototype developed by The Centre devised to filter through and tag Malaysian online discussions - which contain hateful speech.
Speaking at the tracker's virtual launch today, panellist and Institute of Strategic & International Studies (ISIS) senior analyst Tashny Sukumaran contested the idea that cracking down on hate speech could place free speech at risk.
“I think this is a really dangerous misconception that a tracker like #TrackerBenci actually sets out to correct.
“Now, freedom of speech does not mean freedom from objection, right? You do have to understand the consequences of what you say.
“A tracker like this actually could be an invaluable tool in showing people the difference between making a valid criticism and a targeted statement that sets out to demonise a marginalised group.
“The important thing is that we always need more data,” said Tashny.
She noted instances online where discussions on issues such as refugees, women's rights or the #MeToo movement online, there's frequently an influx of hate speech or vitriolic criticism.
“If we don't have a way to categorise this and look at it and objectively study it, you're going to have this idea that everything is alright when it's not,” said Tashny.
Hate speech in democracies
Former Dewan Rakyat speaker Mohamad Ariff Md Yusof – who was a speaker at the same launch – noted the detrimental effects that hate speech has on democracy.
“In my previous job as a Dewan Rakyat speaker, I could see these consequences all too vividly.
"I could see the effects of offensive and hateful speech, whether intentional or not, on the fabric of our democracy and society, even within the confines of Parliament, where MPs address on another as an honourable member can easily descend into unnecessary invectives that border on hate speech - although many will deny it.
“Unfortunately, this is still a facet of our society, which we must minimize all together and eradicate to the extent that it can be done,” said Ariff, adding his hopes of seeing the tracker being made known to parliamentarians.
The #TrackerBenci website features a Benci Calculator – which gauges the likelihood of phrases being hateful - and data gathered from the research.
According to the tracker’s website, 51 percent of words recorded by the Benci Calculator have been identified as potentially harmful.
At present, the tracker is limited to identifying hateful speech found on Twitter in both English and Bahasa Malaysia.
However, The Centre researcher Tham Jia Vern noted the possibility of extending the initiative to other social media platforms in the future.
“We’re looking forward to expanding this initiative to Facebook and beyond, but there are technical limitations for now — for example, getting and analysing their data in a large-scale manner,” she said today.
