As part of the Moore-Sloan Data Science Environment’s ongoing attempt to foster interdisciplinary research, they recently announced the NYU Data Science Seed Grant. The grant is open to NYU faculty researchers, and will fund 6-8 proposals of either up to $25,000, or up to $6,000, in addition to providing software engineering assistance from the data science incubator. Neal Beck, an Affiliated Faculty member at CDS and member of Moore-Sloan’s Working Methods group, is leading the way for this funding opportunity, and answered a few questions below.
How did the NYU Data Science Seed Grant come to be?
Moore-Sloan’s Methods Working group, which plans activities advancing key goals of the Moore-Sloan grant, has always had an interest in promoting interdisciplinary research that begins at the Center for Data Science, and extends to the rest of the University. One particular project we’ve been working on is building out the NYU Data Science Environment, and part of our vision for the environment is to foster collaborations between data scientists, and domain scientists who want to use data science methods as part of their research. We’ve had success in the past, with projects such as the Text Working group, and we want to foster similar types of projects. We decided that a seed grant—matching methods researchers with domain scientists—would be the best way forward.
In addition to the funding itself, the grant offers software engineering assistance from the data science incubator. Can you talk about the assistance that provides?
Some collaborations need engineering assistance on issues such as huge data bases or new graphic user interfaces. For such projects, it is most efficient if collaborations use the infrastructure that has been created by the Moore-Sloan Data Science Environment.
The grant is explicitly for NYU faculty, but could you talk about how graduate students in the Center for Data Science program can get involved?
The NYU Data Science Seed Grant was designed to facilitate collaboration between NYU faculty and domain scientists who are looking to employ data science methods in their research. Faculty collaborators—either on the methods side, or on the domain side—often use a portion of the research money to hire graduate student researchers.
What sorts of projects do you have in mind for the grant?
The critical thing is collaboration between domain and methods faculty. In some areas—such as text or astrophysics—this collaboration is already well developed. In other areas—particularly the field of bio-medical research—the possibilities for synergy are more nascent. The goal of the program is to jumpstart these new method and domain collaborations, and the committee will judge the applications based on how new method and domain collaborations are fostered.
Can you give us a specific example regarding the type of collaborations you are trying to foster?
Even in my own research, I see the need for these collaborations. Personally, my projects involve the use of machine learning to improve standard political science methods. I’m well versed in the field of political science, but to get the maximum mileage out of my research, I want to team up with a methodologist who has a better understanding of machine learning principles. The NYU Data Science Seed Grant, for me, has always been another way of fostering the such collaborations we’re trying to facilitate at the Center for Data Science.
I imagine you have benefitted from research grants at some point in your career. Can you talk about some of the effects these had for you?
As a young researcher, I benefitted from a National Science Foundation grant to work on monetary policy, and I also benefitted from subsequent grants that allowed me and my co-author to follow up with methodological investigations. It is hard to do research without outside funding, and unlike many other seed grant programs, the NYU Data Science Seed Grant was specifically designed to facilitate new collaborations, and not just jumpstart individual research.
Can you talk about why these sorts of collaborations between method scientists and domain scientists are important?
The idea of using the most relevant methods for the most substantive research is fundamental to the field of data science. Methodologists are trained to develop methods, while domain researchers are substantive in the physical, biological or social sciences. The research we want to facilitate requires the joining of these skills. To use the current jargon, we’re trying to break down the silos.
To find out more about the NYU Data Science Seed Grant Funding opportunity, follow the link below: