The KeepA(n)I Project: From basic to applied research and innovation on AI evaluation
Abstract
AI systems can influence the social world, amplifying bias and discrimination. KeepA(n)I, funded by the Cyprus Research and Innovation Foundation (RIF) (“Excellence Hubs” 2022-2024), develops a structured, methodological approach to aid developers in detecting social bias in input datasets and the output of an application. In contrast to existing methods, which often evaluate group and individual fairness in datasets and algorithmic results, KeepA(n)I focuses on the expression of social stereotypes (e.g., based on gender, race, or socio-economic status) and how those are reflected in biases shared by groups of people interacting with the system. Specifically, KeepA(n)I develops a human-in-the-loop approach, methodically exposing stereotypes to reduce negative impact. By engaging humans in the evaluation process through crowdsourcing, KeepA(n)I delivers a diverse (e.g., across cultures) and dynamic (e.g., across contexts and time) evaluation of social norms for machines.
The project focuses on computer vision applications that analyze people-related media (e.g., image content analysis or gender/age recognition from a profile photo) with significant implications for high-risk applications (e.g., screening job applicant profiles, dating or security apps).
In this talk, I will share the “behind-the-scenes” story of how KeepA(n)I was developed, and how it builds directly upon two previous projects that I coordinated (CyCAT, an H2020 Coordination and Support Action; DESCANT, another research project funded by RIF). Thus, key points to be discussed will be:
- How to choose an appropriate call for a given research idea and how to exploit different types of calls to develop one’s research trajectory. I will describe what I gained through each of these projects and how each type of project (CSA, research, innovation) can be beneficial in reaching one’s long-term goals.
- How to strike the right balance in the proposal in terms of the technical details provided.
- How to leverage the synergies between projects and project activities, as to derive the maximum benefit.
In my case, there were periods of time where these projects (and other projects) were overlapping. Thus, active and creative project management is key. I am very happy to answer other questions about my experience (e.g., the differences that I have experienced in terms of managing a European v. national project, how to best communicate with a project officer, etc.), depending on the audience’s interests.
About the Speaker
Jahna Otterbacher received her doctorate from the University of Michigan at Ann Arbor, USA. She is Associate Professor and Vice Dean of the School of Pure and Applied Sciences at the Open University of Cyprus (OUC). At OUC, she leads the Cyprus Center for Algorithmic Transparency (CyCAT), which conducts interdisciplinary research focused on promoting technical and educational solutions for promoting AI transparency and literacy.
Concurrent to this, Jahna co-leads the Fairness and Ethics in AI-Human Interaction (fAIre) group at CYENS, a new center of excellence and innovation in Nicosia, Cyprus, in collaboration with two international Advanced Partners, UCL and MPI. Her research has been funded by the EU’s Horizon 2020 Research and Innovation Program (under Grant Agreements No. 739578 (RISE) & No. 810105 (CyCAT)), as well as the Cyprus Research and Innovation Foundation (under grants EXCELLENCE/0918/0086 (DESCANT) and EXCELLENCE/0421/0360 (KeepA(I)n)). Since 2022, she is included on the Stanford-Elsevier list of the world’s most-cited scholars (in the area of artificial intelligence – image analysis) and in 2023, is one of only five women from Cyprus to have this recognition for her career-long record.