分布式概念格的纵向合并方法研究

Research on vertically combine method of distributed concept lattices

形式背景的核心数据结构概念格在构建过程中具有较高的时间复杂度,这一问题一直是困扰概念格进一步应用的难题。提出了一种在分布式环境下,将形式背景进行纵向拆分,构建出各站点上的概念子格后再进行合并的构建方法。用理论证明了这种合并方法的有效性和完备性,并通过实验表明了该方法与直接用形式背景构造概念格的方法相比,在时间性能上有了显著的改善。

Concept lattice,the core data structure of formal context,has high time complexity when it is constructed.This problem has disturbed the further application of concept lattice.A combine method is developed,which vertically divides the formal context into distributed stations,constructs concept sub-lattices independently,then combines them together.The validity and completeness of this method is proved by theory,and the experimental results show that compared with the algorithm which constructs concept lattice directly from one formal context,this method has dramatically improved the time performance.