Handling partial truth on type-2 similarity-based reasoning

Representation and manipulation of the vague concepts of partially true knowledge in the development of machine intelligence is a wide and challenging field of study. How to extract of approximate facts from vague and partially true statements has drawn significant attention from researchers in the fuzzy information processing. Furthermore, handling uncertainty from this incomplete information has its own necessity. This study theoretically examines a formal method for representing and manipulating partially true knowledge. This method is based on the similarity measure of type-2 fuzzy sets, which are directly used to handle rule uncertainties that type-1 fuzzy sets cannot. The proposed type-2 similarity-based reasoning method is theoretically defined and discussed herein, and the reasoning results are applied to show the usefulness with the comparison of the general fuzzy sets.