Ontology Exploration through Logical Views in Protege

Ontology exploration or modeling becomes a challenging task when knowledge domains are huge and involve many relationships and restrictions. In such circumstances the ability to inspect and modify only specific parts of the model may enable designers to achieve better results, by focusing on limited subsets of the domain. Unfortunately few approaches try to provide explicit solutions to this problem, while the currently available development tools simply provide intelligent collapsing functionalities (e.g., based on User Interest Models) that are supposed to alleviate the problem in some way. The work proposed in this paper extends a visualization plugin for Protege (Ontosphere3d) with so-called "Logical Views" in order to provide an explicit support for visualizing and modeling subsets of a given knowledge domain. We intend a logical view as a user-definable set of ontology entities (concepts and relations) forming an hyper-surface in the multi-dimensional ontology space. Once defined, logical views can be stored inside the ontology in form of annotation properties; as a consequence the view definition is completely independent from the tool employed in its creation and can be easily ported to different platforms and development environments. With respect to current solutions, our approach offers to ontology designers a more flexible instrument that allows an easy and fully customizable partition of the knowledge base into subdomains.