« Back to blog

How Trustworthy are Tweeters? Wellesley College Professors Receive $429k To Find Out

Throughout the day, I’m constantly scanning my Twitter feed, searching for news, tips and advice. It’s Twitter I head to first when I need a story, and I know I’m not alone. With more people turning to social media channels for information, two professors from the Wellesley College Department of Computer Science decided they needed to find a reliable way to determine whether or not the people behind the tweets are trustworthy. Thanks to a grant from the National Science Foundation, now they can.

The Foundation has granted Wellesley $429,000 to develop an application that examines the origin, authenticity and trustworthiness of messages disseminated on social networks to help users make more informed decisions when determining whether or not information can be trusted.

“Users leave a digital trace behind when they make an announcement,” said Eni Mustafaraj, a visiting assistant professor in the Computer Science Department at Wellesley, in a press release. “The application will follow those digital traces to determine whether a message sender is reputable, allowing the user to make a determination about whether a message should be trusted.”

Mustafaraj and Panagiotis Metaxas, a fellow Computer Science professor and founder of Wellesley College’s Media Arts and Sciences Program, made headlines in 2010 when they released research showing how “Twitter-bombing” could manipulate Google real-time results, and how cyber attacks may have affected the Massachusetts senate election, allowing Scott Brown to win over Martha Coakley. Metaxas describes Twitter-bombing as creating a high amount of Twitter accounts and sending a large number — in that case, about 1,000 — tweets within a short period of time.

After the 2010 study, the duo had planned to develop software that would allow users to identify spammers. As more politicians, celebrities, activists and media signed on to Twitter, however, and the site evolved, so did the researchers’ understanding of the platform.

“Spam, itself, is multifaceted and ubiquitous,” Mustafaraj said. “Our focus is more on whether, at any given moment, a user should trust a piece of information that is being broadcasted by a Twitter account.”

Trust will be measured by a combination of several factors, including: past history of the account; whether other Twitter users find the account trustworthy; whether a user has trusted information from the account in the past and have re-tweeted specific messages; or whether an identical piece of information is surfacing from different, independent sources.

“It turns out spam accounts will rank very low, so they will automatically be classified as untrustworthy,” Mustafaraj said.

This application will help social media users identify risks in a world where everyone can be both a producer and a consumer. To help students think about the risks, funding from the grant will also be used to develop an online course for undergraduate and high school students that examines sources and explains what critical thinking actually means in today’s highly interconnected world. Could we soon see a day where we could be saying “so long” to Twitter bots?

from BostInnovation.com http://bostinnovation.com/2011/11/11/how-trustworthy-are-tweeters-wellesley-c...