Peace Ossom joined the MIT Libraries as Director of Research Data Services in August 2024. In this role, she oversees the teams responsible for GIS and Statistical Services, Data Management Services, and Citation Management and Writing Tools. Ossom recently spoke with Bibliotech about the ever-increasing data needs of MIT researchers and partnering with the MIT Office of Research Computing and Data (ORCD) to meet them.

What are some of the most common challenges you see among MIT researchers working with data?

The top challenges MIT researchers face involve access, compliance, and computing. Researchers, particularly those using large language models, need access to large datasets, which can be very difficult to find, obtain permissions for, and share for the purpose of reproducibility. They must first navigate a world of disparate searching platforms, undiscoverable data, restrictive licenses and agreements, and inconsistency in data collection methods and formatting. Regarding compliance, researchers must stay abreast of constantly changing guidelines, policies, and infrastructure here at MIT and from funders and publishers. Lastly, as data needs grow exponentially, researchers often need to utilize a wide variety of platforms while also needing access to robust computing power for the storage and processing of ever-growing datasets.

The MIT Libraries aims to streamline processes for data access, and we advise on next steps for planning and sharing, encourage agreements that allow for greater sharing across the Institute, and utilize standards for consistent data collection and description. We also provide training on data management principles and practices, coding languages, and GIS and data platforms. Last academic year, we had more than 1,000 engagements with MIT community members between individual consultations and workshops.

How do the Libraries work together with ORCD?

MIT researchers have stated the need for more centralized research computing and data support, and ORCD is creating solutions that meet the needs of departments, labs, centers, and institutes (DCLIs) across MIT. ORCD is the Libraries’ close collaborative partner due to the way our services can work together. ORCD provides advanced research computing resources, while the Libraries provides guidance on data management planning, data documentation, data use, and data sharing. Together, we work to develop a robust model of researcher support.

What are some collaborations you’ve worked on recently?

A great example is the joint effort to make Globus data transfer services available to the MIT community. Globus is a system for the secure and efficient transfer of large data, making it a quick and user-friendly solution for those without the computing power. The Libraries and ORCD aim to ensure that the Institute provides not just raw storage and data compliance services, but also robust training and end-to-end lifecycle management, including long-term access and archival support.

We also collaborated with ORCD and the Schwarzman College of Computing on Independent Activities Period (IAP) pilot courses on coding languages, introducing a Practical Computational Thinking series working with advisors in CSAIL and other departments. Anyone from any DLCI can develop practical connections between computing and their home fields. This expands upon the Libraries’ Carpentries@MIT workshop series which began in 2019 and has reached more than 400 participants. We have seen huge demand for these workshops, waitlisting almost as many people as we registered. ORCD staff have become certified Carpentries instructors, enabling us to develop discipline-oriented workshops, including specialized lessons for Biology and Earth, Atmospheric, and Planetary Sciences. This collaboration is helping us to deepen our offerings and meet a clear need across MIT.