User-Friendly pattern-based transformation of OWL ontologies

  • Authors:
  • Ondřej Šváb-Zamazal;Marek Dudáš;Vojtěch Svátek

  • Affiliations:
  • Department of Information and Knowledge Engineering, University of Economics, Prague 3, Czech Republic;Department of Information and Knowledge Engineering, University of Economics, Prague 3, Czech Republic;Department of Information and Knowledge Engineering, University of Economics, Prague 3, Czech Republic

  • Venue:
  • EKAW'12 Proceedings of the 18th international conference on Knowledge Engineering and Knowledge Management
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

The high expressivity of the OWL ontology language often allows to express the same conceptualisation in different ways. A simple example is the difference between 'class-centric' and 'property-centric' modelling styles, such that the same notion is modelled as a class in the former (e.g. 'Purchase') and an object property in the latter (e.g. 'bought from'). Similarly, concept subordination can be expressed via a subclass hierarchy or via individuals connected by a dedicated property (as in SKOS). Such heterogeneity is an obstacle to reusing ontologies in advanced semantic web scenarios. In particular (as mentioned in [1]), two ontologies modelled in different styles are difficult to match or to import into one another. Furthermore, opting for a style when designing an ontology may have an impact on the applicability and performance of reasoners, as some features cause performance problems for certain reasoners. Finally, human users may also prefer viewing ontologies in a certain form, possibly 'folding' parts of their complexity. Semi-automatic transformation of the modelling style of existing ontologies, with the help of tools to be presented in the demo, will alleviate such problems. In the paper we first overview the core framework and then focus on useroriented tools that allow to perform ontology transformation and edit transformation patterns in a friendly way.