中医针灸是国之瑰宝，构建中医针灸专家系统是本研究和实践的目标。在《中医针灸信息库》原系统基础上，结合穴位和疗法的有限可枚举性，提出映射样本穴位疗法为二进制串、一种改进的距离函数和一种通过阈值作为收敛标准的 K-means 聚类算法，为实践目标奠定基础。运用一种评估聚类算法效果设计平衡点的策略，为6万余例针灸治疗案例构建知识库。实践结果表明，该知识库能满足“依病求医”、“依医问病”和知识更新等核心应用需要，查全和查准率好，也能很好地为中医针灸治疗咨询和教学科研提供辅助决策依据。%TCMacupuncture is the national treasure of China,to construct an expert system of TCM acupuncture is the goal of research and practice in this paper.Based on original system of “TCM acupuncture information base ” and in combination with the limited enumerability of acupoints and therapy,we present in the paper to map the therapy of sample acupoints as the binary strings,an improved distance function,and a kind of k-means clustering algorithm with convergence criterion through a tolerance value,and these lay the foundation for the practice target in this paper.We constructed a knowledge library for more than 60 thousand clinic acupuncture treatment cases by applying a strategy of evaluating clustering algorithm effect and designing balance point proposed in this paper.Practice results indicate that the knowledge library satisfies the requirements of core applications,such as “inquiring treatment approaches according to diseases”,“inquiring diseases depending on medical treatment”,and knowledge updating,etc.,with good recall ratio and precision ratio.It can also provide basis of assistant decision making for consulting TCMacupuncture treatment,teaching and science research.
展开▼
