期刊论文

中文

现代信息科技

2096-4706

2023

7

24

115-120+125

10.19850/j.cnki.2096-4706.2023.24.024

20B431:湖南省教育厅资助科研项目 2020JJ4461:湖南省自然科学基金资助项目

本校为第一机构

以郁病辨证为例,开展基于知识图谱的中医智能辅助辨证知识表示与推理研究,提高中医智能辅助辨证模型的构建效率、辨证模型中辨证知识的可视化程度和辨证推理过程的可解释性.以面向智能辅助辨证的郁病辨证知识获取和医案采集工作的成果为基础,构建郁病智能辅助辨证知识图谱,在知识图谱中表示症状知识和证型知识以及二者之间的关系,结合概率推理进行辨证推理测试和分析.构建了刻画 19 种证型和 147 个症状之间关系的郁病智能辅助辨证知识图谱,辨证推理测试获得的初步准确率可达 79.17%、按证型分组统计的准确率最高可达 100%,可根据郁病智能辅助辨证知识图谱对辨证结果进行初步解释.将知识图谱应用...

Taking the syndrome differentiation of depression as an example,this paper carries out research on knowledge representation and reasoning of intelligent assisted syndrome differentiation for Traditional Chinese Medicine(TCM)based on knowledge graph,to improve the construction efficiency of TCM intelligent assisted syndrome differentiation model,the visualization degree of syndrome differentiation knowledge in the syndrome differentiation model,and the interpretability of syndrome differentiation reasoning process.Based on the achievements of knowledge acquisition and medical case collection wo...