基于概念格层面的知识提取方法及应用  

Knowledge level extraction method based on concept lattice and its application

作　　者：毛华[1,2] 刘畅[1,2] 袁晓垒 刘川 MAO Hua;LIU Chang;YUAN Xiaolei;LIU Chuan(College of Mathematics and Information Science,Hebei University,Baoding 071002,Hebei China;Key Laboratory of Machine Learning and Computational Intelligence of Hebei Province,Baoding 071002,Hebei China)

机构地区：[1]河北大学数学与信息科学学院,河北保定071002 [2]河北省机器学习与计算智能重点实验室,河北保定071002

出　　处：《华中科技大学学报（自然科学版）》2024年第11期37-42,92,共7页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金青年科学基金资助项目(12202130);河北省自然科学基金资助项目(A2022201034)。

摘　　要：为解决现有生成概念格算法只能直接输出概念,而不能同时输出每个知识层面及其同一知识层面所蕴含的知识之问题,提出一个将形式背景转化为概念格的知识分层(KL)算法.在宏观上,KL算法实现了将形式背景中的知识可视化;在微观上,KL算法将具体的知识层面及各个知识层面所蕴含的知识一并输出.通过将KL算法与Nextclosure等其他生成概念格算法对比发现:在算法复杂度方面,当形式背景中对象集规模较大时,KL算法在处理数据的速度上明显快于Nextclosure等算法;当形式背景中属性集规模不大时,KL算法与Nextclosure等算法处理数据的速度基本相同;在是否生成Hasse图与知识层面方面,KL算法占有绝对的优势.研究结果表明:KL算法在同一形式背景下提供的结果内容更为丰富,有利于概念格理论应用的推广.To solve the problem that the existing algorithm of generating concept lattice could only output concepts directly,but could not output each knowledge level and the knowledge contained in the same knowledge level meanwhile,a knowledge level(KL)algorithm was proposed to transform formal context into concept lattice.On a macro level,KL algorithm could visualize the knowledge in the formal context.From the micro point of view,KL algorithm could output the specific knowledge level and the knowledge contained in each knowledge level.By comparing KL algorithm with Nextclosure algorithm,it is found that KL algorithm can process data faster than Nextclosure algorithm in terms of complexity when the object set in formal context is large,and when the size of the attribute set in the formal context is not large,KL algorithm has the same data processing speed as Nextclosure algorithm.In terms of generating Hasse graph and knowledge level,KL algorithm has absolute advantage.Research results show that KL algorithm can provide richer results under the same formal context,which is conducive to the popularization of concept lattice theory.

关 键 词：形式概念分析 概念格 知识层面 Hasse图 知识分层算法 

分 类 号：TB182[一般工业技术]