Abstract
As more companies deploy new AI-based systems, there are growing concerns about the explainability of AI from an end-user perspective. Although current explainability techniques, such as SHAP and LIME, improve model transparency, recent studies indicate that they are not always aligned with end-user requirements. In this context, this article explores the close relationship between explainability and interpretability to enable end users to make informed decisions and act confidently. Based on a literature review and an interview-based case study conducted in a supply chain management context, our results highlight the need for functional interpretability for end-users. We also develop the foundations for a theoretical model to reinforce the decision-making process. This study contributes an end-user-centric framework that better aligns with their requirements for complex AI-based deployments, particularly regarding the explainability of AI.
Recommended Citation
EL OUALIDI, Taoufik and Assar, Said, "Extending the Synergy of Explainability and Interpretability in AI-Driven SCM" (2025). MCIS 2025 Proceedings. 1.
https://aisel.aisnet.org/mcis2025/1