Communications of the ACM
Finding Credibility Clues on Twitter
Georgia Institute of Technology researchers have devised a language model that identifies words and phrases that lead to strong or weak perceived levels of credibility on Twitter.
Credit: hackread.com
Researchers at the Georgia Institute of Technology (Georgia Tech) have built a language model that identifies words and phrases that lead to strong or weak perceived levels of credibility on Twitter.
The researchers scanned 66 million tweets linked to nearly 1,400 real-world events, and found the words of millions of people on social media have considerable information about event credibility. They focused on tweets surrounding events in 2014 and 2015, including the emergence of Ebola in West Africa, the Charlie Hebdo attack in Paris, and the death of Eric Garner in New York City.
The researchers asked people to judge the posts on their credibility, and then fed the words into a model that split them into 15 different linguistic categories, each of which included positive and negative emotions, hedges and boosters, and anxiety. The team then examined the words to determine if the tweets were credible or not, and found they matched the humans' opinions about 68% of the time.
"When combined with other signals, such as event topics or structural information, our linguistic result could be an important building block of an automated system," says Georgia Tech professor Eric Gilbert.
The research will be presented in February at the ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW 2017) in Portland, OR.
From Georgia Tech News Center
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA
No entries found