As I mentioned at the start of #blogjune 2014, I’m currently engaged in looking at a new archive management system for the State Records Office of WA. We’ve decided to go with ICA-AtoM for a number of reasons:
- the strong link to international standards, not just for archival description, but also for things like dates (yyyy-mm-dd) and describing the size of digital objects (mebibits);
- the strong community of support for the project, including some fairly significant players like Libraries and Archives Canada;
- the fact that it’s open source; and,
However, I continue to look at the question of different collection management systems and, in particular, the way in which libraries, archives and museums can link data to create better access to their collections in the aggregate. For this reason, the recent discussions on the Collection Management Linkedin group have caught my eye. In a series of posts on Museum IO, Shaun Osborne of Cambridge’s Fitzwilliam Museum has provided an historical overview of collection management systems that will be familiar to archivists and other collection managers, and asks questions about the best way to manage linked data, from questions of the use of RDF (resource description framework) to persistent urls and handles for material in collections. (http://museums-io.org/home/?q=content/museums-and-data-historical-perspective and http://museums-io.org/home/?q=content/museum-data-reality-check-1-rdf). Rupert Shepherd, a museum registrar, has followed up with an impassioned plea for better documentation http://world.museumsprojekte.de/?p=4023) regardless of the management system. As archivists struggle with the ongoing challenge to make their collections more accessible via digitisation, I am reminded of the plea from a researcher to prioritise description before digitisation, and would suggest that this is a debate that we need to make more public and more urgent.
- its ability to export data in a range of formats, that we think will help us take advantage of the semantic web and linked data, and will encourage data reuse from the collections.