Computational provenance: DataONE and implications for cultural heritage institutions

Provenance data is a type of metadata that computer scientists argue can support trustworthy and reliable replication of scientific results. From its origins in scientific workflow systems and database theory, and with concurrent interest from the ecological informatics community, a standard data model (PROV) and extensions for DataONE (ProvONE) have led to initial implementations in several tools commonly used by scientists (R, MATLAB) and in a global federation of scientific data repositories (DataONE). DataONE's support for ingest, storage, indexing and retrieval of provenance data is presented. Implications for libraries are identified. A research agenda for exploring the applicability of the PROV model to cultural heritage institutions — archives and museums — and their digital asset management systems is presented.