中医方剂信息模式学习及方剂聚类

摘要

In this paper,a novel prescription concept model is proposed according to the features of TCM prescriptions information.By utilising the theory of machine learning ,the unstructured text features and it organisation forms of content are learned to derive the extraction pattern of prescription concept.This pattern learning method effectively resolves the poser of the extraction of prescription' s unstructured text,and is used in processing prescription texts for China - US Million Book Digital Library Project.Moreover,to cluster the extracted prescription attribute information exhumes the relationship between the medicine nature and efficacy of different prescriptions, and provides recommending services to Traditional Chinese Medicine prescription enquiry modules.%针对中医药方剂信息的特点提出了一种新型的方剂概念模型,并利用机器学习理论对非结构化的文本特征以及其内容组织形式加以学习,得到方剂概念的抽取模式.该模式学习的方法有效地解决了方剂非结构化文本抽取的难题,被用于中美百万册数字图书馆的方剂文本处理.而对抽取的方剂属性信息作聚类,则发掘了不同方剂药性药效间的联系,为中医药方剂查询模块提供推荐服务.