A semiotics approach to semantic mismatches

Abstract

This paper deals with modeling issues that are common to semiotics and software engineering. A major problem is that vague notions of modeling lead to difficulties in building real-world representations for use in the software development life cycle. Specifically, this paper focuses on building a consistent representation that eliminates semantic mismatches through analysis of semiotics. Semantic mismatches occur when the same term is associated with multiple concepts. Semiotics provides a good theoretical foundation for UML research, since a UML diagram can be considered a sign made up of signs. This paper introduces a new approach based on the so-called Flowthing Model (FM) to represent semiotics notions of sign, interpretant, and object, for use in studying the problem of semantic mismatches. The conclusion is that diagrammatic UML representation can lead to the appearance of such problems and that FM description provides separate streams such that no mixing can occur among terms as things that flow.