Semi-Automatic Generation of a Software Testing Lightweight Ontology from a Glossary Based on the ONTO6 Methodology
Guntis Arnicāns, Dainis Romans, Uldis Straujums · Frontiers in artificial intelligence and applications · 2013
We propose a methodology of semi-automatic obtaining of a lightweight ontology for software testing domain based on the glossary “Standard glossary of terms used in Software Testing” created by ISTQB. From the same glossary many ontologies might be developed depending on the strategy for extracting concepts, categorizing them, and determining hierarchical and some other relationships. Initially we use the ONTO6 methodology that allows identification of the most important aspects in the given domain. These identified aspects serve as the most general concepts in taxonomy (roots in the concept hierarchy). By applying natural language processing techniques and analyzing the discovered relations between concepts, an intermediate representation of lightweight ontology is created. Afterwards the lightweight ontology is exported to OWL format, stored in the ontology editor Protégé, and analyzed and refined by OWLGrEd – an UML style graphical editor for OWL that interoperates with Protégé. The obtained lightweight ontology might be useful for building up a heavyweight software testing ontology.