Research Objects for Audio Processing: Capturing Semantics for Reproducibility

Earlier work has identified the potential for reuse and reproducibility when applying workflow systems to audio analysis and Music Information Retrieval. In this paper we extend this approach with the introduction of Research Objects to capture semantic information about the use of workflows within the audio research and development process. Once aggregated, the metadata encapsulated in a Research Object can be used to manage and disseminate research output, providing a well structured foundation for meeting the needs of reproducibility. We report on the development and deployment of a software suite that practically applies this notion of Research Objects to capture the semantics surrounding the use of an audio processing workflow, and reflect upon how this might be further integrated with lower level semantics from the audio processing domain.