Masterarbeit MSTR-2023-78

Li, Zhixian: Enabling Integration between MUSE4Anything and OWL-Based Ontology Tools.
Universität Stuttgart, Fakultät Informatik, Elektrotechnik und Informationstechnik, Masterarbeit Nr. 78 (2023).
53 Seiten, englisch.

Ontologies are used for knowledge representation and information management in various domains, e.g., semantic web or digital humanities. They contain the elements of a domain and how they are related. For example, using ontologies, researchers in digital humanities can model different music pieces and use the ontology to identify recurring patterns. Especially for the research of using ontologies in digital humanities to identify patterns, the ontology tool MUSE4Anything emerged from the MUSE and MUSE4Music projects, offering a more generic ontology model. However, using and editing ontologies in different tools might be beneficial if the tools offer different features for the ontology modeler, e.g., the digital humanities researcher. Nevertheless, many ontology tools represent their ontologies using a different metamodel and format. Therefore, ontologies modeled in the MUSE4Anything tool are not directly usable by another ontology tool, such as Protégé. Over the past decades, the Web Ontology Language (OWL) has emerged as a well-established ontology representation, which is, for example, used in Protégé. To tackle the problem of interoperability between MUSE4Anything and OWL-based ontology tools, we propose a mapping of MUSE4Anything ontologies to the OWL format in this thesis. To map the formats, we first derived the JSON-schema-based metamodel of MUSE4Anything ontologies by analyzing MUSE4Anything’s OpenAPI specification and database models. We implement an export to OWL function in MUSE4Anything using our proposed mapping. We apply the template language concept to parse MUSE4Anything ontologies to the XML-based OWL format. This way, ontologies created with MUSE4Anything can be opened and edited by OWL-based ontology tools like Protégé. This enables researchers to apply a more extensive toolset to work with MUSE4Anything ontologies. To evaluate our work, we mapped two MUSE4Anything ontologies to the OWL format and validated the interoperability using Protégé as a representative for OWL-based tools.

Abteilung(en)Universität Stuttgart, Institut für Architektur von Anwendungssystemen
BetreuerLeymann, Prof. Frank; Bühler, Fabian
Eingabedatum20. Februar 2024
   Publ. Institut   Publ. Informatik