中医典籍知识图谱

本产品通过先进的文本挖掘和自然语言处理技术，对1100多部中医古籍中记载的有效个案进行病证理法方药靶的条文分类、实体的识别和概念术语的归一化处理。建立了中药成分、靶点、药物、疾病之间的相互关系，并与TCMID2、TCMSP、DrugBank、HERB、MedDra等公开数据库概念映射。基于知识图谱实现了经典古方智能检索和推荐系统，提高专科专病验方、有效成分、副作用等的挖掘和筛选效率。

This product adopts advanced text mining and natural language processing technologies to conduct classification of relevant clauses, entity recognition and normalization processing of conceptual terms for effective clinical cases recorded in more than 1,100 ancient Chinese medical classics, covering symptoms, syndromes, pathogenesis, treatment principles, prescriptions, medicinals and targets. It establishes the interrelationships among traditional Chinese medicine (TCM) ingredients, targets, medicinals and diseases, and performs concept mapping with public databases including TCMID2, TCMSP, DrugBank, HERB and MedDra. Based on knowledge graphs, it has developed an intelligent retrieval and recommendation system for classical ancient prescriptions, and improved the efficiency of mining and screening targeted prescriptions for specific departments and diseases, active ingredients, side effects and other related contents.

搜集汇总

背景与挑战

背景概述

中医典籍知识图谱数据集基于1100多部中医古籍，通过文本挖掘和自然语言处理技术构建，包含中药成分、靶点、药物、疾病等实体及其相互关系，并与多个公开数据库映射，支持经典古方的智能检索和推荐。