GDMKGLM: A Knowledge Graph Dataset for Gestational Diabetes Mellitus

The GDMKGLM Dataset is a comprehensive, structured collection of medical knowledge resources specifically curated for research on Gestational Diabetes Mellitus (GDM). This dataset was constructed to support the development and evaluation of the GDMKGLM system, a Graph Retrieval-Augmented Generation (GraphRAG) based question-answering system described in our publication. The dataset integrates diverse medical knowledge sources including clinical guidelines, research literature, medical textbooks, and patient education materials, transformed into a structured knowledge graph containing 973 medical entities and 854 semantic relationships.