Translations between RDF(S) and Conceptual Graphs

Abstract

Though similarities between the Semantic Web language RDF(S) and languages of the Conceptual Graphs family have often been pointed out, the differences between these formalisms have been a source of difficulties while trying to translate objets of a language into the other. In this paper, we present two such transformations, that have been implemented into the CoGUI platform, and discuss their respective strengths and weaknesses.

A Generic Platform for Ontological Query Answering

“…Please refer to the paper [16] for a discussion on the relevance of this dataset. Using those generated knowledge bases then requires an initial translation from RDF into first order logic expressions (done offline, according to [3]). Please note that the initial RDF translation step has not been taken into account when reporting on storage times.…”

Graph-based ontology reasoning for formal verification of BREEAM rules

“…This transformation is achieved through the following principles (Baget, Croitoru, Gutierrez, Leclère, & Mugnier, 2010): the acknowledgement of the distinction between the basic component of an ontology with the translation of classes into concept types, properties into binary relations, and instances into individual markers; the preservation of the visual appeal and formal meaning of conceptual graphs; the clear differentiation between ontology and data.…”

“…Furthermore, efforts should be made to improve the output of the automated regulations to enhance the generation of human readable documentation in compliance checking processes. The linking of the graph configuration with the semantic web and Table 4 Correspondences between RDF, Conceptual Graphs and logic (Baget et al, 2010).…”

Learning Commonalities in RDF

“…the problem we study: the (minimal) cover graph-based lgg of tree-shaped RDF graphs is clearly a forest-shaped RDF graph in general. In CGs, the so-called simple CGs with unary and binary relations correspond to particular RDF graphs (e.g., a property URI in a triple cannot be the subject or object of another triple, a class -URI or blank node -in a τ triple cannot be the subject of another τ triple nor the subject or object of another non-τ triple, etc), which may feature the four RDFS constraints, and for which RDF entailment is limited to the use of these RDFS contraints only [10]. In these equivalent RDF and CG fragments, we may interchangeably compute lggs with the CG technique in [11] or ours.…”