证候是中医辨证的主要依据，为了量化分析中医内科诊疗规律，本项目优化知识图谱技术，通过构建领域本体，将中医内科诊疗信息拆分为病、证和证候要素，以图谱的方式可视化存储其“症-病-证”的内容与关系，并通过知识计算对四诊信息进行辨病与辨证，探索中医复杂系统分析。. 本项目拟采集中医内科病证相关信息，结合前期积累构建领域本体，拟采用BiLSTM+CRF模型进行命名实体识别，R-BERT模型进行关系抽取，特征向量提取+相似度计算的方案构建实体链接，使用Neo4j数据库存储知识图谱并嵌入设计一种“症-病-证”相关性识别算法，实现对各个病证关键证候要素的提取，通过改良K-BERT或TransE模型实现基于知识图谱的病证识别与推理方法。. 本项目旨在发现和分析中医内科诊疗过程中关键证候要素及其作用，为挖掘中医内科诊疗规律、提高中医疗效、实现中医现代化和智能辅助辨证提供理论和方法依据。

Syndrome is the main basis for differentiation of syndromes in traditional Chinese medicine., in order to quantify and analyze the rules of diagnosis and treatment in internal medicine of TCM , this research will optimize the knowledge graph technology, divide the diagnosis and treatment information in internal medicine of TCM into disease, syndrome types and syndrome elements, store and visualization the content and relationship of "symptom-disease-syndrome type" in the graph. Through knowledge calculation, the four diagnostic information of TCM can be distinguished and differentiated to realize disease and syndrome identification, so as to explore the complex system analysis of TCM.. This research plans to collect relevant information in internal diseases of TCM , build domain ontology in combination with our previous research works, use BiLSTM+CRF model to identify named entities, R-BERT model to extract relationships, feature vector extraction and similarity calculation to build entity links, use secondary Neo4j database to store knowledge graph and embed and design a "symptom-disease-syndrome type" correlation identification algorithm to extract key syndrome elements of disease, and improve K-BERT or TransE model to realize identification and reasoning methods in internal diseases of TCM based on the knowledge graph.. The purpose of this research is to discover and analyze the key syndrome elements and their functions in the diagnosis and treatment in internal diseases of TCM, and provide theoretical and methodological basis for mining the rules of diagnosis and treatment in Disease and Syndrome in Internal Medicine of TCM, improving the curative effect of TCM, and realizing the modernization of TCM and Intelligence-assisted distinguish symptoms of diseases in TCM.