Leveraging Representation Learning for the Construction and Application of a Knowledge Graph for Traditional Chinese Medicine: Framework Development Study.

Experiments showed that the clinical KG representation learning and application framework is effective for knowledge discovery and decision-making assistance in diagnosis and treatment. Our framework shows superiority of application prospects in tasks such as KG-fused multimodal information diagnosis, KGE-based text classification, and knowledge inference-based medical question answering.