Acupoint recommendation and knowledge graph analysis based on improved convolutional network

Personalized acupoint recommendation: The paper addresses the challenge of making precise acupoint recommendations tailored to specific symptoms while accounting for symptom-acupuncture relationships.

TCM data scarcity: The paper identifies scarcity of traditional Chinese medicine data as a limitation for conventional recommendation approaches.

Personalized acupoint recommendation: The paper addresses the challenge of making precise acupoint recommendations tailored to specific symptoms while accounting for symptom-acupuncture relationships.

TCM data scarcity: The paper identifies scarcity of traditional Chinese medicine data as a limitation for conventional recommendation approaches.

Property-fusion graph convolutional network: The paper proposes an improved graph convolution network with property fusion, named PEGCN, for acupoint recommendation.

Acupoint attribute enrichment: The method extracts attribute information of acupuncture points to enrich their representations.

Attention-enhanced GCN relation modeling: The approach combines a GCN architecture with a focus system to model relations between symptoms and acupuncture points.

CiteSpace bibliometric knowledge mapping: The paper uses CiteSpace 6.3 to perform bibliometric analysis and generate knowledge maps for keywords, authors, and organizations.

Property-fusion graph convolutional network: The paper proposes an improved graph convolution network with property fusion, named PEGCN, for acupoint recommendation.

Symptom-acupoint relation capture:

Bibliometric knowledge map generated:

Symptom-acupoint relation capture: