Soft Ontologies as Fuzzy RDF Statements

Soft ontologies enable knowledge representation in a flexible way because they are more susceptible to changes in their interrelationships over time. The transformation of a nonstandardized model of information into a computer-interpretable representation is not an easy task. Our investigation expands the concept of soft ontologies to fuzzy data represented as fuzzy RDF triples. In this paper, we elaborate a process that transforms a soft ontology implemented in the form of a matrix of probabilities into a fuzzy RDF dataset. The study presents the way the matrix of probabilities are used in the representation and how the data elements are triplified exploring fuzzy characteristics. We apply the proposal in an experimental scenario, by constructing a soft ontology which expresses a repertoire of actions in an mBot robot. Then, the facts are triplified to fuzzy RDF statements. Our results present original aspects related to the transformation of soft ontologies to fuzzy RDF triples.

I. Introduction

Data categorization is essential to establish a comprehensible relation between humans and machines. This is a key factor to improve the capacities of new interactive systems. For this purpose, an ontology is a way to represent knowledge making use of vocabulary definitions which describes objects and their relationships in a certain domain [1].