基于概念提升的概念格更新构造算法  

An Updating Algorithm of Concept Lattice Based on Concept Promoting

作　　者：杜秋香[1] 张继福[1] 张素兰[1] 

机构地区：[1]太原科技大学计算机科学与技术学院,太原030006

出　　处：《太原科技大学学报》2009年第1期1-6,共6页Journal of Taiyuan University of Science and Technology

基　　金：山西省自然科学基金(2006011041)

摘　　要：概念格因其结构生动简洁地表明概念之间的泛化-例化关系,成为一种有效的数据挖掘工具。然而直接从形式背景中直接构造概念格的效率较低,为了提高概念格的构造效率,针对形式背景中的多个属性合并,给出了一种基于概念提升的概念格更新构造算法UCP.该算法充分利用了属性合并之前的概念格,仅对部分概念结点进行更新处理,从而提高了概念格的构造效率。最后,以恒星光谱数据作为形式背景,采用VC++6.0和O rac le9 i为开发工具,实验验证了该算法的有效性。Concept lattice is an effective tool for data mining since its structure reflects the vivid simplicity of the relationship between generalization and specialization among concept lattices. But the efficiency is lower when the concept lattice generates directly from the formal context. For the merging of many attributes in the formal context, an updating algorithm UCP of concept lattice based on concept promoting is presented in order to improve the construction efficiency. The algorithm makes full use of the concept lattice before the attributes merge. Some parts of nodes are updated, therefore the updating efficiency is improved. Finally, experiment results show the validity of the algorithm by using VC ＋ ＋ 6.0 and Oracle9i as development tools and taking the celestial spectrum data as the formal context.

关 键 词：概念格 概念提升 更新 UCP算法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]