Professor of Information Systems, W. P. Carey School of Business, Arizona State University
Closing the Representation Loop: Building A Design Theory For Interpretable Algorithmic Representations
Abstract: Algorithmic interpretability is necessary to build trust, ensure fairness, and track accountability. However, there is little theoretical work that helps guide the design of interpretable algorithmic representations (IAR). In this paper, we aim to build a design theory for IARs. Using representation theory as the kernel theory, we propose the key meta-requirements for generating interpretable representations of algorithms. We then discuss potential meta-designs and testable hypotheses of IARs using various modeling grammars. Finally, we illustrate the design of IARs through a wide range of algorithms such as recommender systems and LLMs.
Bio: Dr. Hong Guo is a Professor of Information Systems at Arizona State University. Hong studies emerging IS phenomena (such as digital platforms, digital games, algorithmic interpretability, net neutrality, and business data visualization) and firms’ corresponding strategies. Hong’s research has been published in top business journals such as MIS Quarterly, Information Systems Research, Manufacturing & Service Operations Management, and Production and Operations Management. She currently serves as a senior editor for Production and Operations Management. She also served as an associate editor for Information Systems Research between 2022-2023 and MIS Quarterly between 2017-2020.