We on the OTTER project are gradually working our way through various issues and challenges surrounding implementation of our project. I have written elsewhere about the challenges of developing evaluation criteria. Another matter of concern to me is the question of “metadata”. The importance of this piece of information to data retrieval goes perhaps as far back as the days of the library catalogue. The term is generally understood to mean “data about data”. Stated differently it refers to a structured piece of information about another item of information. This added piece of information, called metadata i.e. beyond or behind the data, is seen as bringing the added benefit of increased accessibility, usability tracking, migration of data objects to successive platforms and providing evidence for system improvement. Up to this point metadata is meaningful to me. However, beyond this point one begins to wander into a world of “meta data standards” with different emphases and foci such as:
- Dublin Core (DC)
- Text Encoding Initiative (TEI)
- Encoded Archival Description (EAD)
- Data Documentation Initiative (DDI)
Whilst these standards are universally recognized, their applicability within a particular context remains a matter of confusion. The extent to which these standards evolve to cater for the ever increasing types of digital resources is also unclear. For the OTTER team, we have been trying to answer two basic questions regarding metadata:
- Which metadata standards should we apply to our Learning Objects (LO) to make them OERable?
- What level of granularity is acceptable to make the OER accessible and repurposable?
- How much is enough and how much is too much?
I have read elsewhere that the JISC has decided to take a new approach to metadata where rather than mandate standard metadata application profile; the decision is left to projects to make for flexibility. This to me is a reason way to resolve the current dilemma. What needs to be guarded against though is a situation where “flexibility” will compromise “accessibility”. If this happens there would be more metadata matters to resolve.
Samuel Nikoi (2 July 2009)