Statistics-Based Graphical Modeling Support for Ontologies

Andreas Emrich; Frieder Ganz; Dirk Werth; Peter Loos

In: Petr Dini (Hrsg.). The Seventh International Conference on Advances in Semantic Processing. International Conference on Advances in Semantic Processing (SEMAPRO-13), located at NexTech 2013, September 29 - October 4, Porto, Portugal, IARIA, 9/2013.


Models are a conceptualization and aperture of the real world. The asynchronous characteristics of this model construction poses a significant problem especially in highly dynamic and evolving environments. Hence, models need to be permanently checked against the data they represent. From this new challenges for modeling tools arise: Contemporary modeling tools must be able to anticipate environmental events and changes and to provide appropriate support for knowledge engineers. This paper presents a conceptual approach to process events, collect usage statistics and leverage this information for automatic and semi-automatic modeling support of ontologies. In a prototypical evaluation, a plugin for ProtegŽe is developed, that allows for editing ontologies and visualizing new insights according to captured statistics. The paper concludes with two distinct show cases for business process modelling and sensor networks.

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence