演繹オブジェクト指向データベースを用いた 演奏生成アプリケーションオントロジー

平田 圭二, 平賀 瑠美

Application Ontology for Performance Rendering based on Deductive Object-Oriented Database Approach

Keiji Hirata, Rumi Hiraga

Abstract:
This paper discusses the roles of a musical ontology for performance rendering application. Basically, performance rendering consists of two stages: score analysis and the application of a performance model to the result of the score analysis. However, few attempts have successfully programmed these two stages to realize natural and expressive performance. This is because of the characteristics of music: partial, tacit, incomplete, and subjective. It turned out that almost all conventional systems employed ad hoc methods with heuristics, although a few exceptional systems worked well in very limited musical genres. We think that an essential problem here is the lack of evaluation-improvement cycle of system development. We propose the introduction of a musical ontology to build a performance rendering system. Since the musical ontology allows us to treat music in a formal and explicit way, it will enable musical concepts to be real entities in a computer program. With the musical ontology, we can elaborate on cause-and-effect relationships among the system output, the score analysis, and the performance model; furthermore, a user will gain more system controllability. In this paper, we first clarify domain ontology for music proposed so far, then describe an application ontology for performance rendering based on the deductive object-oriented database technique.