基于形式概念分析的证名域方法的研究

发布时间：2018-02-21 00:56

本文关键词： 中医辨证形式概念分析属性偏序分层优化知识表示与发现证名生成　出处：《燕山大学》2012年硕士论文　论文类型：学位论文

【摘要】：辨证论治不仅是中医学理论的核心，也是中医诊断学的特点与和优势。中医辨证的客观与否直接影响临床疗效。要做到客观的辨证仅仅依靠文献知识和专家的经验还不够，还要通过对大量的数据挖掘分析，从中找出辨证的规律。致力于中医辨证研究的学者们都在努力的寻找能够科学处理那些具有不完整性和不确定性的病情资料和数据的有效途径，并做了许多探索性的工作。至今还无法用一种标准化的数学语言对中医辨证论治原理进行一个科学详尽的描述。形式概念分析的理论以数学化的概念和概念层次为基础，可以用来解决证和证素构成网络的描述性问题。由于中医中的证型和证素的个数比较多，如果利用传统概念格来描述证（对象）与证素（属性）之间的关系，则根据概念层次结构生成的Hasse图中层与层之间存在着较多的交叉连线，无法直观的从杂乱无章的图中发现有用的中医诊断知识。为了解决这个问题Hasse中层与层之间交叉数较多的问题，本文提出了一种基于形式概念分析的属性偏序结构图的生成方法，并将这种方法与基于行列变换的分层优化方法相结合，构造出的唯一的属性偏序结构图具有概念层次关系简练，无交叉连线，类别意义清楚，层次关系鲜明的特点。根据层次关系可视化表示，标注对象与属性注解，建立起的基于属性偏序关系的知识表示和发现方法，有助于中医辨证诊断知识体系的建立和生成证名的检索。
[Abstract]:Syndrome differentiation and treatment is not only the core of TCM theory, but also the characteristics and advantages of TCM diagnostics. The objectivity of TCM syndrome differentiation directly affects the clinical efficacy. It is not enough to rely only on literature knowledge and expert experience to achieve objective syndrome differentiation. And by analyzing a lot of data mining, To find out the law of syndrome differentiation. Scholars working on TCM syndrome differentiation are trying to find an effective way to deal with the incomplete and uncertain disease information and data. So far it is impossible to use a standardized mathematical language to describe the principle of TCM dialectical treatment in a scientific and detailed way. The theory of formal conceptual analysis is based on the mathematical concepts and conceptual levels, and can be used to solve the descriptive problem of the network of syndromes and syndromes. Because of the large number of syndromes and syndromes in traditional Chinese medicine, If Ge Lai, a traditional concept, is used to describe the relationship between evidence (object) and element (attribute), there are many cross lines between middle layer and layer of Hasse graph generated according to concept hierarchy. It is impossible to intuitively find useful diagnostic knowledge of traditional Chinese medicine (TCM) from random maps. In order to solve this problem, a method of generating attribute partial order structure graph based on formal concept analysis is proposed in this paper. By combining this method with the hierarchical optimization method based on column and column transformation, the unique attribute partial order structure diagram is constructed, which has a concise concept hierarchy relationship, no cross line, and clear category meaning. According to the visual representation of hierarchical relations, annotating objects and attributes, a knowledge representation and discovery method based on attribute partial order relation is established. It is helpful for the establishment of knowledge system of TCM syndrome differentiation and the retrieval of syndromes.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R241;R311

【相似文献】