Sponsored message
Audience-funded nonprofit news
radio tower icon laist logo
Next Up:
0:00
0:00
Subscribe
  • Listen Now Playing Listen
AirTalk

Can AI Spot Trolls Online? These Caltech Researchers Think So

This photo taken on January 7, 2010 shows a woman typing on the keyboard of her laptop computer in Beijing. China declared its Internet "open" on January 14 but defended censorship that has prompted Web giant Google to threaten to pull out of the country, sparking a potential new irritant in China-US relations. China employs a vast system of Web censorship dubbed the "Great Firewall of China" that blocks content such as political dissent, pornography and other information viewed as objectionable and the issue looks likely to shape up as the latest addition to a growing list of disputes between China and the United States over trade, climate change and human rights.     AFP PHOTO / Frederic J. BROWN (Photo credit should read FREDERIC J. BROWN/AFP/Getty Images)
A new study out of Caltech created a proof of concept project that was able to use machine learning to identify harassment online.
(
FREDERIC J. BROWN/AFP/Getty Images
)
Listen 22:28
Can AI Spot Trolls Online? These Caltech Researchers Think So

For algorithms, spotting trolling or harassment online largely relies on monitoring keywords -- a blunt instrument that doesn’t account for context or the evolution of conversations. But a new machine learning algorithm might change that. 

A new study out of Caltech, "Finding Social Media Trolls: Dynamic Keyword Selection Methods for Rapidly-Evolving Online Debates” created a proof of concept project that was able to use machine learning to identify harassment online, using “word embedding models” that are better at understanding context. 

We take a closer look at this new method and its promise.

Guests:

Anima Anandkumar, professor of computing and mathematical sciences at Caltech; one of the principal investigators on the study  

Michael Alvarez, professor of political science at Caltech; he is one of the principal investigators on the study