电子学报 ›› 2023, Vol. 51 ›› Issue (1): 11-17.DOI: 10.12263/DZXB.20211559

• 学术论文 • 上一篇    下一篇

带对象结构信息形式背景的概念知识发现与演化

闫梦宇1,2, 李金海1,2, 刘文奇1,2, 张文修3   

  1. 1.昆明理工大学数据科学研究中心, 云南 昆明 650500
    2.昆明理工大学理学院, 云南 昆明 650500
    3.西安交通大学数学与统计学院, 陕西 西安 710049
  • 收稿日期:2021-11-22 修回日期:2022-06-30 出版日期:2023-01-25
    • 通讯作者:
    • 李金海
    • 作者简介:
    • 闫梦宇 女,1994年5月出生于山东省枣庄市.现为昆明理工大学系统科学专业博士研究生.主要研究方向为形式概念分析、复杂网络、概念认知学习等.E-mail: yanmengyu2016@163.com
      李金海(通讯作者) 男,1984年1月出生于江西省上饶市.2012年毕业于西安交通大学数学与统计学院.现为昆明理工大学数据科学研究中心教授、博士生导师.主要研究方向为概念格、粗糙集、粒计算、概念认知学习等.
    • 基金资助:
    • 国家自然科学基金 (11971211)

Conceptual Knowledge Discovery and Evolution in Formal Context with Object Structure Information

YAN Meng-yu1,2, LI Jin-hai1,2, LIU Wen-qi1,2, ZHANG Wen-xiu3   

  1. 1.Data Science Research Center, Kunming University of Science and Technology, Kunming, Yunnan 650500, China
    2.Faculty of Science, Kunming University of Science and Technology, Kunming, Yunnan 650500, China
    3.School of Mathematics and Statistics, Xi’an Jiaotong University, Xi’an, Shaanxi 710049, China
  • Received:2021-11-22 Revised:2022-06-30 Online:2023-01-25 Published:2023-02-23
    • Corresponding author:
    • LI Jin-hai
    • Supported by:
    • National Natural Science Foundation of China (11971211)

摘要:

形式概念分析是数据处理和知识发现的有力工具,然而仅利用形式概念分析中的传统方法已不能满足大数据环境的需求,这是因为大量数据不仅包括对象-属性信息,还有对象与对象之间的结构信息,因此扩展形式概念分析进行数据处理与知识发现是一个重要的研究课题.基于此,本文定义了带对象结构信息形式背景,并在形式概念的基础上引入连通性提出全局结构信息形式概念和局部结构信息形式概念(统称为结构信息形式概念),进而讨论基于结构信息形式概念的知识发现问题;由于现实中数据的动态变化及信息的传播与扩散是不可避免的,又进一步探讨了当带对象结构信息形式背景对象结构矩阵或连接矩阵发生变化时结构信息形式概念的演化规律,根据不同情况分别给出了对应的算法并对它们的时间复杂度进行分析;此外,通过实验表明了结构信息形式概念动态更新方法的可行性与有效性.

关键词: 形式概念分析, 结构信息形式概念, 知识发现, 概念演化, 动态更新

Abstract:

Formal concept analysis is a powerful tool for data processing and knowledge discovery. However, only using the traditional methods in formal concept analysis cannot meet the needs of big data environment since most data not only contain object-attribute information but also the structure information between objects. As a result, it is an important research topic to extend formal concept analysis for data processing and knowledge discovery. Based on the above analysis, this paper puts forward the notion of a formal context with object structure information. And then, the global structural information and local structural information formal concepts are proposed (collectively called structural information formal concepts) by combining with connectivity. Meanwhile, the problem of knowledge discovery on the basis of structural information formal concepts is discussed. Considering the fact that the dynamic changes of data and the dissemination or diffusion of information in our daily life are inevitable, we further explore the evolution rules of structural information formal concepts when a formal context with object structure information is updated gradually. Algorithms are developed for updating different structural information formal concepts when object structure matrix or connection matrix changes, and their time complexity are also analyzed. In addition, we conduct some experiments to show the feasibility and effectiveness of the proposed structural information formal concept updating methods.

Key words: formal concept analysis, structural information formal concept, knowledge discovery, concept evolution, dynamic updating

中图分类号: