Ying Ding
There is a lot of power in data, and if that data is analyzed and understood, it can change the world and make a real impact on lives.
Bringing together data scientists and social scientists was the goal of the Data Science Meets Social Science Symposium hosted at Woodburn Hall at Indiana University on Nov. 3. The event showcased collaborations between data science at SoIC and social science faculty in the College of Arts and Science on campus, and explored the possibilities of combining the approaches of each field.
“Data science can help social science scrape large scale datasets and provide help with big data analytics,” says Ying Ding, an associate professor of informatics and an organizer of the event. “Social scientists can help data scientists identify good research questions to work on. It’s a great collaboration between two of the strengths on our campus.”
For instance, a collaboration between Yong-Yeol Ahn, an assistant professor of informatics and computing, and Fabio Rojas, an assistant professor in the department of sociology at IU, explored how network framework allows hidden patterns to be revealed in social and cultural data. Using network data, Ahn and Rojas found they could chart the general mood of the country based on such diverse areas as the optimism of song lyrics and the tone of Twitter comments throughout the day. That data can be used to discover how networks are organized and what they can reveal.
David Wild, an associate professor of informatics and computing and the director of the SoIC data science program, presented a talk that suggested using healthcare data analysis to find evidence paths that can help researchers better understand the system-wide impact of drugs on the human body. Can the existing data identify precision interventions with existing drugs or other chemicals for particular patient groups? The emerging data model blends pharmaceutical research, healthcare delivery, and public health initiatives, and finding evidence paths through data is the first step toward discovery of other treatments.
David Crandall, an associate professor of informatics and computing, has long used images culled from social media sites such as Flickr to create 3D models of locations, and the geotagging of digital photos allows him to track social connections and mobility. The same data also can be used to document environmental changes at a certain location and track those changes over time, and providing visual proof of those changes could help better understanding of climate change.
Inna Kouper, the assistant director of the SoIC Data to Insight Center, presented a talk about the HathiTrust Research, which provides an opportunity for large-scale text analysis to unlock insights into social science. Rojas also gave a talk exploring whether the glass ceiling for women exists in social media discussions. Alex Hollingsworth, an assistant professor at the IU School of Public and Environmental Affairs, examined the relationship between economic conditions and opioid deaths. Using data from every death from 1999-2014, he found a robust relationship between an increase in the unemployment rate and the rate of fatal and non-fatal opioid-related visits to the emergency room. Hollingsworth admitted a lot of different forces could be in play, but it didn’t change the clear relationship between the two data points and suggested the relation be a component of future policies designed to tackle opioid abuse.
“The Data Science Meets Social Science Symposium really showcases the groundbreaking work that can be done when you are innovative with your thinking,” says Rick Van Kooten, IU’s vice provost for research. “The collaborative spirit shown in such an effort between SoIC and COAS takes the strengths of our campus and builds creative and important partnerships with real-world impacts.”
The success of this year’s event, which featured a packed room of attendees and a waitlist, has inspired plans for future events.
“We organized this event to bring people together, and we look forward to expanding it,” Ding says. “After talking with Alex, we also might want to organize a roundtable event with SPEA faculty to create consulting projects with our Data Science Consulting Lab. It was a great event that showed the power of collaboration.”