G-OWL: A Complete Visual Syntax for OWL 2 Modeling and Communication

Tracking #: 2550-3764

This paper is currently under review
Gilbert Paquette
Michel Héon

Responsible editor: 
Aldo Gangemi

Submission type: 
Full Paper
Semantic web ontologies are usually modeled using standard text-based syntaxes such as OWL/XML, Functional, Manchester or Turtle. Over recent years, there has been an increasing need for representing ontologies visually to help ontological engineers or modelers represent elicited knowledge from domain expert, big data, model data structures or simply present data schemas and metadata to general users. We believe a visual representation is an essential way for understanding knowledge and to help elaborate formal ontologies for their communication and their use by humans. In this paper, we present the Graphical Ontology Web Language (G-OWL), a visual syntax for the graphical modeling and visualization of OWL 2 or RDFS ontologies. In line with previous research in cognitive science, G-OWL uses syntactic and semantic principles that simplify both its use and its interpretation. Indeed, the use of typology and polymorphism makes it possible to minimize the number of visual signs in a grammar, thus reducing the cognitive load on users, while preserving the formal character of the ontology. This G-OWL visual syntax is integrated in a software tool called OntoCASE4G-OWL to support the elaboration of ontologies and their translation to standard text-based syntaxes such as Turtle. This paper aims to present the definition of the G-OWL visual syntax and to demonstrate its highly readable character through an objective assessment of criteria such as: semiotic clarity, semantic transparency and graphic complexity. The G-OWL visual syntax will also be compared with other visual syntaxes and will be evaluated in order to measure its highly human-readability in reading activities, modeling or new knowledge deduction by novice, intermediate and expert in ontology modeling.
Full PDF Version: 
Under Review