摘要
利用聚类思想对概念格进行压缩。根据概念之间外延和内涵的差别得到概念之间的相似度;从概念之间的相似度出发,基于谱系聚类思想对所有概念进行聚类。根据聚类的结果,利用K删除变换对概念格进行压缩并且证明了概念格经过压缩之后得到的仍然是一个完备格。
This paper proposes a method to compress a concept lattice using clustering. Firstly,the similarity measure between formal concepts is obtained through the extent and the intent,and then,the concepts are clustered by hierarchical clustering. Finally,a K-deletion transformation is used to realize the compression of concept lattice and it is proposed that the compressed one is also a concept lattice.
出处
《陕西理工学院学报（自然科学版）》
2016年第3期78-82,共5页
Journal of Shananxi University of Technology:Natural Science Edition
基金
宝鸡文理学院硕士启动项目(ZK16114)
关键词
形式背景
概念格
谱系聚类
压缩
相似度
formal context
concept lattice
hierarchical clustering
compression
similarity degree
