Tufts Universitys Office of the Vice Provost for Research hosted its 10th Research Day on May 2, focusing on the emerging field of data science a topic that looks to analyze enormous amounts of data collected by modern technology, such as sequencing techniques, electronic medical records and social media outlets.
"Now, thinking about data science is almost like how we used to think about reading and writing," Professor of Civil and Environmental Engineering Shafiqul Islam said. "How do we get insight? Most of the data is noise. The challenge then for academic communities is to figure out ways to get the signal from the noise."
Associate Professor in the Department of Computer Science Soha Hassoun, chair of the Research Day Committee, proposed data science as this year's topic because the field is expanding and she feels that Tufts should join the dialogue surrounding the subject.
Hassoun explained that the committee had three objectives for the Research Day. First, it wanted to identify what problems are being worked on in data science at Tufts, what techniques are being used and what data sets could be shared. Next, the committee wanted to identify people with different skills who can collaborate on big data issues. Lastly, it aimed to start a conversation at Tufts to discuss enhancing infrastructure for research and education in data science.
"There are lots of people from different disciplines talking about data science," Hassoun said. "Harvard, MIT, Columbia, University of Virginia and other schools are already ahead of us in their various initiatives, and it would be really good for us as a community [to see] what we can do with our resources."
The day was broken up into three rounds of lightning talks. The first session, "Big Data Analysis: From Fundamentals to Engineering Applications," provided attendees with background information about data science, according to the Office of the Vice Provost for Research. Session two, "Data-Driven Discoveries," examined how researchers use large data sets and what they are discovering within those sets. The third session, "Data Enabling Medicine," explored medical applications for large data.
"[The Research Day] features three very different areas to give you a flavor of what's going on here at Tufts," Barbara Booras, project administrator in the Office of the Vice Provost for Research, said. "Broadly, we are using the day as an educational tool to educate about resources that we have already available to faculty and students related to data."
The lunchtime digital poster session gave graduate and Ph.D. candidates the opportunity to share their research in areas of data science. According to the Office of the Vice Provost for Research, the goal of the Research Day was mainly to educate the Tufts community about the available resources connected to data. Tisch Library, for example, had a table during the poster session to answer questions and talk about its data management plan that is currently in place for students and faculty.
"So many times faculty say, 'We want to collaborate across campuses but how do we facilitate those collaborations?'" Booras said. "This is an opportunity to bring people together to start a conversation that can continue outside of the event, as well."
Sophomore Mary O'Kane, a quantitative economics major, was one of approximately 10 undergraduate students who attended the event. O'Kane said she appreciated hearing about all the opportunities to get involved with Tufts research in the field of data science through the poster session, but wished that more undergraduates had been able to attend.
"Listening to what the graduate and Ph.D. candidates are doing was really cool," she said. "It would have been great if Tufts had scheduled it at a different time. If it wasn't during finals period ... [they] probably would have gotten a better undergraduate response."
The day concluded with a panel presentation that focused on the infrastructural and educational importance of continuing to develop data science at Tufts. A discussion and question and answer session followed, in which participants and attendees debated how Tufts could best improve its data science courses.
Some panelists, including Professor of Computer Science Carla Brodley, advocated for the strategic hiring of more data science researchers. Others, like Professor of Civil and Environmental Engineering and Associate Dean for Research Elena Naumova, argued that the priority should be pooling Tufts' preexisting resources in order to broaden course offerings.
"We don't have a statistics department," Naumova said. "I feel that's a tremendous limitation. I feel our students have very limited access to high level courses in statistics. We need to put more resources into building a wide range of courses our students can take."
Namouva also emphasized the importance of communicating across various departments and levels of the university. Islam echoed Namouva's sentiments.
"I think the first thing we need to do is to pull together the resources we currently have," Islam said. "Once we do that, we can identify the knowledge gaps. Then we can decide what can be done to fill these gaps."
O'Kane agreed.
"One thing that stood out to me that was brought up was the lack of cohesiveness between the departments," she said, reflecting on the panel discussion. "[We] just ... [need to] have an awareness of what people are doing and utilize [that] expertise. They talked about needing to hire a lot of people, and a lot of it comes down to simple infrastructure issues and management."12
