New Methods for Recognizing and Resolving Contradictions in Knowledge

Problem statement: When analyzing and processing external information by SEMS modules, it may be necessary to consider knowledge that contains contradictions. Previously, the authors proposed methods for dealing with paradoxes by means of E-structures as well as some methods for resolving contradictions in knowledge bases by adding a new attribute. The problem is that those results do not explain the reasons for many other paradoxes and some contradictions in knowledge. Purpose of research: Continue the search and justification of new methods for resolving contradictions and explaining paradoxes. Results: A new model of polysyllogistics-like reasoning is proposed on the basis of the laws of algebra of sets, in which the shortcomings of E-structures are corrected. New methods for analyzing and resolving contradictions are developed. These include, in particular, the replacement of a universal proposition with a particular one in contradictory reasoning and the identification of binary relations for which the self-reference leads to a paradox. Practical significance: The results obtained make it possible to furnish the knowledge base of SEMS modules with new methods for recognizing and resolving contradictions when analyzing incoming information in the form of knowledge and reasoning.