�山ǿ��

When thousands of demonstrators filled Egypt’s Tahrir Square in 2011, they carried powerful revolutionary tools—smartphones. Now, a �山ǿ�� Mississauga professor is using technology to make it easier to collect and organize large amounts of data gleaned from social media about protest events.

joined UTM’s as an assistant professor in July 2016. With a background in computer science, math and sociology, she has focused on using computer-aided data collection to analyze how social movements begin and grow.

“I study how participants in social movements have used social media to create mobilizing behaviours, like organizing a protest, getting people to join a protest or broadcasting information about protest events,” she says.

One of Hanna’s major projects considers the role social media played in the development of the Arab Spring movement. The 2011 protests were thought to have initially mobilized through social media platforms. Hanna analyzed posts from a 10,000 member Facebook group that formed in 2008, looking for information about the movement’s early days. “I collected messages from the group’s first two months of the group’s existence to see how people were mobilizing and communicating with each other,” she says. “A lot of messages were focused on coordinating events, with information about meet-ups and locations. But I also saw many messages of support, along with messages that seemed to indicate involvement and engagement from diaspora outside of Egypt who wanted to get involved.”

Hanna relies on large-scale data collections and computational tools for her research, and says one of the challenges is that the process of reading and hand coding information gleaned from thousands of social media posts can be expensive, time-consuming and labor-intensive. To make data collection faster and simpler, .

This experience has directly influenced her most recent project—the creation of the Machine-learning Protest Event Data System (MPEDS), a computer program that parses text for information about protest events. With natural language programming, MPEDS can be taught to extract information about dates, times, participants and more from publications with a variety of writing styles.

“I am working to identify algorithms that would be useful to help identify protest events and to extract those elements of protest events that are relevant to social movement scholarship,” Hanna says. “For example, it is useful to know the location, date and form of protest, along with the issue at hand, the target of the protest, size and the movement and organizations involved in the event.” The collected information can help sociologists track how protests arise, and to monitor the subsequent success or failure of those movements.

Hanna hopes to launch MPEDS in fall 2016 with a preliminary data set on black protest movements, a project she is working on with University of Wisconsin-Madison sociologist Pamela Oliver.