Research on attribute granular computing and its fuzzification method based on qualitative mapping

Abstract

A novel attribute fuzzy granule cluster based on qualitative transformation mechanism of attribute and the computing model thereof was put forward. This paper first establishes information granules belonging to the attribute topological space by qualitative mapping, then figures out expression of attribute cluster operators under various granularities to form a hierarchical attribute granule simplex, studies how to determine granularity for fuzzy information system and finds out a method for generating fuzzy attribute granule. The result proves that the method can better model overall attribute analysis ability of human by formalization, so that the computer will better simulate such ability, providing a new research approach of granular computing method for artificial intelligence. © 2011 Springer-Verlag Berlin Heidelberg.