基于核方法的模糊对象语言概念格聚类分析  

Clustering Analysis of Linguistic Concept Lattice with Fuzzy Object Based on Kernel Method

作　　者：庞阔 康宁 邹丽 鲁明羽[1] 郑宏亮[3] PANG Kuo;KANG Ning;ZOU Li;LU Ming-yu;ZHENG Hong-liang(School of Information Science and Technology,Dalian Maritime University,Dalian 116026,China;School of Computer Science and Technology»Shandong Jianzhu University,Jinan 250102,China;School of Computer and Information Technology,Liaoning Normal University,Dalian 116029,China)

机构地区：[1]大连海事大学信息科学技术学院,辽宁大连116026 [2]山东建筑大学计算机科学与技术学院,山东济南250102 [3]辽宁师范大学计算机与信息技术学院,辽宁大连116029

出　　处：《模糊系统与数学》2022年第1期31-46,共16页Fuzzy Systems and Mathematics

摘　　要：在现实生活中,存在着大量语言值数据。为了解决在语言环境中不确定信息的聚类问题,本文提出了一种新的机器学习方法,即基于核方法的模糊对象语言概念格聚类分析模型。该模型通过融合层次聚类与概念格聚类的原理,在寻找到层次聚类局部最优层次的同时优化概念格聚类中的概念选择与概念构造问题。具体地,提出模糊对象语言概念格及其相关性质,它的外延与内涵分别用模糊集与语言术语集来描述,不仅可以表达模糊信息和语言值信息,而且可以描述不同群体之间的语言信息差异;分别分析层次聚类与模糊对象语言概念格聚类,在此基础上,结合核方法计算模糊对象语言概念格中样本间的相似程度,构造基于层次聚类方法的模糊对象语言概念格聚类分析模型;实验结果表明,该模型是有效的,可以集成层次聚类与概念格聚类的优势。In real life,there are a lot of linguistic value data.In order to solve the clustering problem of uncertain information in linguistic value environment,this paper proposes a new machine learning method,namely linguistic concept lattice with fuzzy object clustering analysis model based on kernel method.By combining the principle of hierarchical clustering and concept lattice clustering,the model not only finds the local optimal level of hierarchical clustering,but also optimizes the problem of concept selection and concept construction in concept lattice clustering.Specifically,the linguistic concept lattice with fuzzy object and its related properties are proposed.Its extension and intension are respectively described by fuzzy sets and linguistic term sets.It can not only express fuzzy information and linguistic value information,but also describe the differences of linguistic value information among different groups.The hierarchical clustering and linguistic concept lattice with fuzzy object clustering are analyzed respectively.On this basis,the similarity between samples in linguistic concept lattice with fuzzy object is calculated by combining kernel method,and the linguistic concept lattice with fuzzy object clustering analysis model based on hierarchical clustering method is constructed.Experimental results show that the model is effective and can integrate the advantages of hierarchical clustering and concept lattice clustering.

关 键 词：模糊对象语言概念格 核方法 层次聚类 语言术语集 形式概念分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]