Answers are grounded in this paper and cite their sources.
Explainable Artificial Intelligence (XAI) aims to make the reasoning processes of AI models transparent and interpretable, particularly in complex decision-making environments. In project-based sectors such as construction, where AI-based decision support systems (DSSs) are increasingly deployed, limited attention has been given to the role of evidence in supporting the reliability, transparency, and accountability of their outputs. Without explicitly incorporating evidence into the design of XAI-enabled DSSs, the validity of explanations and the trustworthiness of system recommendations remain open to question. To address this challenge, our paper presents a theoretical, evidence-based means–end framework developed through a narrative review. The framework provides an epistemic foundation for designing XAI-enabled DSSs that can produce Meaningful Human Explanations (MHEs) aligned with end-users' knowledge needs and decision-making contexts. It emphasizes evaluating the strength, value, and utility of different forms of evidence used to support AI-generated explanations. Although the framework is developed with construction professionals as its primary end-users, it also applies to other stakeholders, including developers, regulators, and project managers, whose epistemic goals may differ. By explicitly integrating evidence into the design of XAI-enabled DSSs, the framework provides a conceptual basis for evidence-based, transparent, and accountable AI-supported decision-making. Given its theoretical foundation and broad applicability, it can inform the design of evidence-aware XAI systems across a range of engineering and industrial domains.
No takes yet. Share an insight, caveat, or question.
Love et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: