We propose a Trustworthy Fairness Metric and its measurement methodology to evaluate the fairness of different AI-based solutions proposed by the actors utilizing their trust during decision-making in the Food-Energy-Water (FEW) sectors. Since the standardization of the trustworthiness of AI systems is fundamental, the proposed metric is a compelling advance in this process, whereas other approaches stay at the high-level principles. Trust management system is the basis of the measurement methodology as it incorporates human involvement. This metric captures and quantifies the fairness of the solutions evaluated by the actors having different views and is illustrated in decision-making scenarios generated by AI for FEW sectors. Also, the metric and its measurement methodology can be conveniently adapted to various fields of AI. We present that our metric successfully captures the fairness of solutions in multi-stakeholder decision-making.
CITATION STYLE
Uslu, S., Kaur, D., Rivera, S. J., Durresi, A., Durresi, M., & Babbar-Sebens, M. (2022). Trustworthy Fairness Metric Applied to AI-Based Decisions in Food-Energy-Water. In Lecture Notes in Networks and Systems (Vol. 450 LNNS, pp. 433–445). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-99587-4_37
Mendeley helps you to discover research relevant for your work.