集值序信息系统的信息熵和知识粒度

摘要

针对已有文献中二元优势关系定义过于宽松的不足，在集值序信息系统中结合对象间的不同优势程度，提出δ-优势关系的概念；基于δ-优势关系，将信息熵和知识粒度引入集值序信息系统中进行不确定性的度量。结论表明提出的信息熵和知识粒度可以精确地度量集值序信息系统的不确定性。%To overcome the shortcoming of the existing dominance relations, a new δ-dominance relation that considers the superiority degree between objects is proposed for set-valued ordered information systems. And then, based onδ-domi-nance relation, the concepts of information entropy and knowledge granulation are introduced in set-valued ordered infor-mation systems to measure the uncertainty. Results show that information entropy and knowledge granulation can evaluate the uncertainty of knowledge in set-valued ordered information systems.