形式概念分析在不完备信息系统中的知识获取  

Knowledge Acquisition in Incomplete Information System Based on Formal Concept Analysis

作　　者：李想[1] 王素格[2,3] 李德玉[2,3] 康向平[4] 翟岩慧[2,3] 

机构地区：[1]山西大学数学科学学院,太原030006 [2]山西大学计算机与信息技术学院,太原030006 [3]山西大学计算智能与中文信息处理教育部重点实验室,太原030006 [4]同济大学电子与信息工程学院,上海201804

出　　处：《计算机科学》2014年第7期250-253,260,共5页Computer Science

基　　金：国家自然科学基金资助项目(61175067;61272095;60970014);山西省科技攻关项目(20110321027-02);山西省回国留学人员科研项目(2013-014)资助

摘　　要：在实际应用中,信息系统通常不满足完备性,直接采用经典粗糙集无法进行有效求解。针对该问题,在粗糙集中引入形式概念分析,通过探讨二者之间的关系,在不完备信息系统中提出了两大理论融合下的知识获取模型。在该模型中,首先将不完备信息系统转换为单值形式背景,提出了相容概念和相容概念格,然后基于相容概念对不完备信息系统中的一些常见问题(如上下近似算子、核、约简等)的求解进行了研究,最后探讨了相容概念在不完备决策表中的应用。该模型不仅对两种理论之间的融合研究进行了有益探索,而且为不完备信息系统中的一些基本问题求解提供了一种新的思路。In practical applications, some common problems in information system can＇t be solved effectively based on the classical rough set due to its incompleteness. To solve this problem, the paper introduced formal concept analysis in- to rough set, proposed a knowledge acquisition model in incomplete information systems by discussing the relationship between formal concept analysis and rough set theory. First by converting incomplete information system into a one-va- lued context, consistent concepts and consistent concept lattice were proposed. Then some common problems in incom- plete information systems were studied, such as upper and lower approximations, cores, reducts, etc. Finally, the applica- tion of consistent concepts in a decision table was discussed. The proposed model not only explores the fusion of two theories greatly, but provides a new idea for solving some basic problems of incomplete information system.

关 键 词：形式概念分析 粗糙集 相容关系 不完备信息系统 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]