Graduate students spent the summer performing data science research in marine biology, public health, public transportation, digital security, and more.

 

Thanks to data science graduate students and a public service program that matches students with a common-good project, marine biologists will be able to estimate the size and health of a whale solely from a photograph. When whales are encountered near the surface of the ocean, it can be useful to quickly assess their size for conservation, evaluation, and other purposes. Chhandak Bagchi and Gizem Cicekli, computer science master’s students at the Robert and Donna Manning College of Information and Computer Sciences, spent 12 weeks researching and developing a tool to automatically estimate the size of a whale from an aerial image. Working with partners from the UMass Amherst biology department, the Woods Hole Oceanographic Institution and Aarhus University, they developed a prototype that, once validated, may be deployed on marine research ships to process whale images taken from drones in real time.

 

Bagchi and Cicekli were research fellows in Data Science for the Common Good (DS4CG), a summer program operated by the Center for Data Science (CDS) that trains aspiring data scientists to work on real-world problems that benefit the common good. Teams of computer science master’s students collaborate with nonprofits and government agencies to apply computational solutions to “big data” problems. Students work directly with the partner agencies, as well as with CDS staff and professional data science advisors from CDS’s pool of corporate affiliates. Last summer, six DS4CG teams worked with partners including the Massachusetts Department of Public Health, the Pioneer Valley Transit Authority, the World Resources Institute, and the Center for Digital Resilience, on projects that serve the public good.

 

Not only do DS4CG students gain experience applying data science to real-world problems, they also have the opportunity to submit papers and workshops for major conferences. Collin Giguere and Sowmya Vasuki Jallepalli, DS4CG fellows on the Georeferencing of Historical Imagery project, developed an image-processing system that uses ground control points to automatically georeference historical aerial images. The tool can be used to reveal changes in land and waterways over time. Their project was accepted as a panel presentation at the Northeast Arc User Group’s (NEARC) Fall 2021 conference, an annual gathering of users of Geographic Information System (GIS) software from Environmental Systems Research Institute. 

 

DS4CG will be seeking nonprofit organizations and government agencies to partner with for the 2022 program. Visit our website to learn more. 

For more information about last summer’s projects, visit the 2021 DS4CG Projects page.