Digital Long-Term Preservation, Provenance and Linked Open Data
Sam Coppens, Erik Mannens or Raf Vandesande, with contribution by Rik Van de Walle
Nowadays, the need for digital long-term preservation is growing. A lot of the digital information produced merely a decade ago is in danger of getting lost as technologies are changing. This also threatens a lot of information from heritage institutions in Belgium. Analogue carriers are degrading and digitisation is a must, but digital copies are also subjected to many risks, e.g., data carrier corruption, or file formats becoming obsolete. For this, a distributed long-term preservation platform is being developed overcoming those risks, assuring the accessibility of the information. This distributed archive harvests the data records from the institutions, disseminates the data as Linked Open Data, and archives the data for the long-term by providing and tracking its provenance information. This provenance information is described using PREMIS OWL, our developed and (LOC) approved semantic implementation of the PREMIS 2.0 data dictionary. The developed platform also enables the provenance information on the Web. For this, the long-term preservation platform relies on datetime content negotiation, an extension to the HTTP protocol, proposed by Memento, and a special HTTP link header for discovery of the provenance information. The datetime content negotiation will deliver the right version of the data to the end-user, the HTTP provenance link header points to the provenance information of that version. This provenance information links the current version of the data to its other versions.
elag2011_Coppens.m4v (46,05 MB)