加权概念格及其渐进式构造  

Weighted Concept Lattice and Incremental Construction

作　　者：张继福[1] 张素兰[2] 郑链[1] 

机构地区：[1]北京理工大学机电工程学院,北京100081 [2]太原科技大学计算机科学与技术学院,太原030024

出　　处：《模式识别与人工智能》2005年第2期171-176,共6页Pattern Recognition and Artificial Intelligence

基　　金：国家863高技术研究发展计划基金(No.2003AA133060)

摘　　要：概念格是数据分析与知识提取的有效工具。为了充分利用其进行数据分析和知识提取,本文通过对概念格的内涵引入权值,提出一种加权概念格,拓广了概念格的结构。同时由用户设立概念格内涵的最小阈值,构造了一种频繁加权概念格,并由此给出了频繁加权概念格的构造算法及其分析。最后,通过实例说明其是有效可行的。Concept lattice is an efficient formal tool for data analysis and knowledge extraction. The presented researches presume that all the intensions of the concept lattice are equally important. However, the importance of the attributes consisting of the intension is often different in practice. In order to take advantage of concept lattice for analyzing data and knowledge extraction completely, we introduce the weight value into the intension of the concept lattice and present a new concept lattice, a weighted concept lattice (WCL), which develops the concept lattice structure. The lowest intension threshold is defined by the user. According to the threshold, we build a frequent weighted concept lattice (FWCL) and give one incremental construction algorithm and algorithm analysis for the FWCL. Finally, the example illustrates its efficiency and feasibility.

关 键 词：概念格 加权概念格 频繁加权概念 加权内涵 渐进式构造算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]