A Retrieval-Augmented Generation enhanced expert system framework for supply chain risk management: a healthcare industry application

Ren, J ORCID: 0000-0003-4640-824X, Wang, L, Liangrokapart, J and Kritchanchai, D (2026) A Retrieval-Augmented Generation enhanced expert system framework for supply chain risk management: a healthcare industry application. Enterprise Information Systems. ISSN 1751-7575

Preview

Text A Retrieval-Augmented Generation enhanced expert system framework for supply chain risk management a healthcare industry application.pdf - Published Version Available under License Creative Commons Attribution. Download (2MB) | Preview

Abstract

Traditional expert systems struggle to manage high-stakes healthcare supply chain risks, such as drug shortages and counterfeiting, since knowledge bases require manual updating. This paper proposes a framework combining Retrieval-Augmented Generation (RAG), fuzzy inference, and Dempster-Shafer evidential reasoning to extract and structure risk knowledge from unstructured documents. A UK and China hospital case study identifies drug shortage and counterfeiting as the highest-priority risks. A small pilot of the RAG component alone shows grounded extraction outperforming a no-retrieval baseline, with no hallucination flagged. To our knowledge, no earlier framework uses RAG for upstream knowledge acquisition in supply chain risk management.

Item Type:	Article
Uncontrolled Keywords:	0806 Information Systems; 1503 Business and Management; Business & Management; 3503 Business systems in context; 3507 Strategy, management and organisational behaviour; 4609 Information systems
Subjects:	H Social Sciences > HE Transportation and Communications H Social Sciences > HF Commerce > HF5001 Business > HF5410 Marketing. Distribution of Products T Technology > T Technology (General)
Divisions:	Engineering and Built Environment
Publisher:	Taylor & Francis
Date of acceptance:	13 October 2026
Date of first compliant Open Access:	20 October 2026
Date Deposited:	20 Oct 2026 07:50
Last Modified:	20 Oct 2026 07:50
DOI or ID number:	10.1080/17517575.2026.2750200
URI:	https://researchonline.ljmu.ac.uk/id/eprint/29709

View Item