City Research Online

Abstract

This thesis addresses the representation of and reasoning on musical knowledge in the Semantic Web. The Semantic Web is an evolving extension of the World Wide Web that aims at describing information that is distributed on the web in a machine-processable form. Existing approaches to modelling musical knowledge in the context of the Semantic Web have focused on metadata. The description of musical content and reasoning as well as integration of content descriptions and metadata are yet open challenges. This thesis discusses the possibilities of representing musical knowledge in the Web Ontology Language (OWL) focusing on chord sequence representation and presents and evaluates a newly developed solution.

The solution consists of two main components. Ontological modelling patterns for musical entities such as notes and chords are introduced in the (MEO) ontology. A sequence pattern language and ontology (SEQ) has been developed that can express patterns in a form resembling regular expressions. As MEO and SEQ patterns both rewrite to OWL they can be combined freely. Reasoning tasks such as instance classification, retrieval and pattern subsumption are then executable by standard Semantic Web reasoners. The expressiveness of SEQ has been studied, in particular in relation to grammars.

The complexity of reasoning on SEQ patterns has been studied theoretically and empirically, and optimisation methods have been developed. There is still great potential for improvement if specific reasoning algorithms were developed to exploit the sequential structure, but the development of such algorithms is outside the scope of this thesis.

MEO and SEQ have also been evaluated in several musicological scenarios. It is shown how patterns that are characteristic of musical styles can be expressed and chord sequence data can be classified, demonstrating the use of the language in web retrieval and as integration layer for different chord patterns and corpora. Furthermore, possibilities of using SEQ patterns for harmonic analysis are explored using grammars for harmony; both a hybrid system and a translation of limited context-free grammars into SEQ patterns have been developed. Finally, a distributed scenario is evaluated where SEQ and MEO are used in connection with DBpedia, following the Linked Data approach. The results show that applications are already possible and will benefit in the future from improved quality and compatibility of data sources as the Semantic Web evolves.

Publication Type:	Thesis (Doctoral)
Subjects:	M Music and Books on Music > M Music Q Science > QA Mathematics > QA75 Electronic computers. Computer science Z Bibliography. Library Science. Information Resources > ZA Information resources > ZA4050 Electronic information resources
Departments:	Doctoral Theses School of Science & Technology > Computer Science School of Science & Technology > School of Science & Technology Doctoral Theses

Export

Downloads