SemioSem: A semiotic-based similarity measure

Abstract

This paper introduces a new similarity measure called SEMIOSEM. The first originality of this measure, which is defined in the context of a semiotic-based approach, is to consider the three dimensions of the conceptualization underlying a domain ontology: the intension (i.e. the properties used to define the concepts), the extension (i.e. the instances of the concepts) and the expression (i.e. the terms used to denote both the concepts and the instances). Thus, SEMIOSEM aims at aggregating and improving existing extensional-based and intensional-based measures, with an original expressional one. The second originality of this measure is to be context-sensitive, and in particular user-sensitive. Indeed, SEMIOSEM is based on multiple informations sources: (1) a textual corpus, validated by the end-user,whichmust reflect the domain underlying the ontology which is considered, (2) a set of instances known by the end-user, (3) an ontology enriched with the perception of the end-user on how each property associated to a concept c is important for defining c and (4) the emotional state of the end-user. The importance of each source can be modulated according to the context of use and SEMIOSEM remains valid even if one of the source is missing. This makes our measure more flexible, more robust and more close to the end-user's judgment than the other similarity measures which are usually only based on one aspect of a conceptualization and never take the end-user's perceptions and purposes into account. © Springer-Verlag 2009.