AUTHOR=Li Xiaoran , Han Xiaosong , Wei Siqing , Liang Yanchun , Guan Renchu 
  
TITLE=Acupuncture and tuina knowledge graph with prompt learning
  
JOURNAL=Frontiers in Big Data
  
VOLUME=Volume 7 - 2024
  
YEAR=2024
  
URL=https://www.frontiersin.org/journals/big-data/articles/10.3389/fdata.2024.1346958
  
DOI=10.3389/fdata.2024.1346958
  
ISSN=2624-909X
  
ABSTRACT=Acupuncture and tuina, acknowledged as ancient and highly efficacious therapeutic modalities within the domain of Traditional Chinese Medicine (TCM), have provided pragmatic treatment pathways for numerous patients. Given the recent global focus on Chinese medicine, both Chinese and international scholars have increasingly emphasized its significance. To address the scarcity of structured and quantitative evaluation frameworks within the realm of acupuncture and tuina, this paper embarked on the construction of a comprehensive domain knowledge repository.Concurrently, this study introduced a Named Entity Recognition (NER) method bolstered by prompt learning. This approach was finely tuned to accommodate the intricacies of Chinese textual content. The results showcased an F1 score of 42.68% on Chinese text, surpassing the baseline model by a notable 2%, thereby yielding superior outcomes. Subsequently, the proposed methodology synergized with a Trie tree-based extraction technique to establish a precisionenhanced entity extraction model. This model laid the groundwork for a rule-based relationship extraction framework amenable to contemporary datasets. Consequently, a pragmatic and pertinent knowledge graph pertaining to ancient Chinese acupuncture and tuina was instantiated. This comprehensive graph encompassed a cohort of 10,346 entities interlinked by 40,919 relationships.