Authored by Amy Latessa (OoR Advanced Research Computing team) and Mark Chalmers (UC Libraries – Science and Engineering Libraries)
On November 5th, UC Libraries with collaboration from the OoR Advanced Research Computing team, hosted the 6th annual UC Data Day. The theme of this year’s virtual event was Bias, Miscommunication, and Equity in Data and featured a Keynote by Heather Krause, two interactive panel sessions, and a weeklong virtual screening of the documentary film Coded Bias. This year’s event had 145 attendees and 29 views of the film.
Following the opening remarks of Xuemao Wang, Vice Provost for Digital Scholarship and Dean and University Librarian of UC Libraries, Data Day commenced with the keynote by Heather Krause, PStat, and Founder of We All Count. Heather challenged the traditional notion of data’s objectivity, reminding attendees that researchers must make choices and these choices are not objective. Heather also walked participants through several examples of research questions and demonstrated how the questions could be better designed with equity in mind and the onus of change on systems. She emphasized to participants that there is no “silver bullet” against bias, prejudice, and injustice, and we all must remain vigilant and consider who’s lived experiences are being centered, even at the onset of a project such as when formulating a research question. Heather was engaging and took many thought-provoking questions from the audience.
Heather stayed as the moderator of the first panel session on Bias, Miscommunication, and Equity in Data. She was joined by Dr. Ebony Ruhland (Assistant Professor in the School of Criminal Justice at the University of Cincinnati), Dr. Jeff Blevins (Professor of Journalism at the University of Cincinnati), and Dr. Margarita Boenig-Liptsin (Research Fellow at the Paris Institute of Advanced Study). The panel discussed communication and ethics of data framing it around the term coined by Dr. Blevins, “the social media marketplace of ideas”.
The film Coded Bias, which informed the second panel discussion, follows M.I.T. Mediba Lab computer scientist Joy Buolamwini along with data scientists, mathematicians, and global watchdog groups as they expose the prevalent discrimination within algorithms. It demonstrates that the very machine-learning algorithms intended to avoid prejudice are only as unbiased as the humans and historical data programming them.
The plot of the film served as the theme for the final interactive panel discussion moderated by Chris Collins (Founder and Technical Lead of the UC Center for Simulations & virtual Environments Research). Chris was joined by Dr. Anissa Tanweer (Research Scientist at the University of Washington’s eScience Institute), Dr. Deeptankar DeMazumder (Director of the Artificial Intelligence and Precision Medicine Center of Excellence, and PI of the Neurocardiology Research Lab in the Division of Cardiology) and Nirmalya Thakur (Instructor and researcher with the Department of Electrical Engineering and Computer Science at the University of Cincinnati). These panelists outlined the “good vs. bad of AI” discussing the pros and cons of the ABC’s: Availability of data, Braided systems, and Computational power and how AI understands user interactions. Dr. Demazumder highlighted how AI facilitates information but is not an advocate of AI making the decision. This concept of the importance of AI literacy was echoed by the other panelists.
We invite you to visit the Data Day website, review the sessions and continue the dialogue at AskData@uc.edu or attend an upcoming Data and Computational Science series event.