A team of University professors are researching the impacts social media has on society, with the help of a $4 million research grant.

Tarek Abdelzaher, professor in Computer Science, along with his team, will begin creating models and systems allowing them to analyze the results and make predictions about social media impact.

“We received this $4 million grant to study how information and misinformation spreads on social media and how it changes the way people think,” Abdelzaher said in an email.

The team is in preliminary stages of work as they received the grant in late 2017 from the Defense Advanced Research Projects Agency. According to their website, DARPA hopes to lead researches that focus on “transforming revolutionary concepts and even seeming impossibilities into practical capabilities.”

With the team’s research headed towards analyzing and solving social media patterns, it can help empower individuals with knowledge about social media impacts, Abdelzaher said. He compares it to spam filters in email inboxes.

“On social media, the problem is a bit more insidious because sometime you do not even realize that something you heard was fake,” Abdelzaher said. “That’s how rumors end up spreading and bad information gets perceived as truth. We need a new generation of ‘spam filters’ for social media and our team is part of the solution.”

Members of the team also include professors in Engingeering David M. Nicol and Jiawei Han. Both hope to make way for social media research with the help of the grant as well.

“Social media will play a very important role in our future social life. That’s why we actually want to study a lot of very different models for social media, how they provocate or what things likely may happen,” Han said. “You can predict something and you may want to intervene, or doing something that may be good for society.”

The team of professionals hopes to build its research models in order to collect data on social media and its impacts. They also hope to devise a system to predict future social media patterns using the results.

“We shall first build models that understand how information spreads on social media, then [we] shall then use those tools to build a simulator – a machine that helps predict what’s going to happen on social media next,” Abdelzaher said. “With that machine, we can see undesirable or dangerous things before they happen and can identify fake sources and bad information before it causes harm.”

While Abdelzaher is particularly focused on the results of the research, both Han and Nicol are focused on the models which they are working on for the research.

“My concern has to do with developing the methodology, ties, setbacks if you will, of being able to assess how these models are. Because if you have a model that is not particularly accurate, not particularly predictive, as configured to the way things actually are in a real network,” Nicol said.

[email protected]