Learning to Comprehend and Trust Artificial Intelligence Outcomes: A Conceptual Explainable AI Evaluation Framework
Peter E.D. Love, Jane Matthews, Weili Fang, S.J. Porter, Hanbin Luo, Lieyun Ding · IEEE Engineering Management Review · 2023
Explainable artificial intelligence (XAI) is a burgeoning concept. It is gaining prominence as an approach to better understanding how AI solutions' outputs can improve decision-making. Evaluation frameworks to enable organizations to understand XAI's what, why, how, and when are yet to be developed. Thus, we aim to fill this void by developing a conceptualcontent,context,process,andoutcome(CCPO) evaluation framework to justify XAI's adoption and effective management using construction organizations as a backdrop for the paper's setting. After introducing and describing the proposed novel CCPO framework for operationalizing XAI, we discuss its implications for future research. The contributions of our paper are twofold: (1) it highlights the need for organizations to embrace and enact XAI so that decision-makers and stakeholders can better understandwhyandhowa specific prediction materializes; and (2) it provides a frame of reference for organizations to realize the business value and benefits of XAI.