There are widely recognized metadata standards available to describe digital content. Each standard usually targets the use of a specific category of resources in a specific context. Relying on standards avoids reinventing the wheel while benefiting from best practices developed by experts. Moreover, standards enable interoperability (between the systems that support the standards). Finally, adopting widely accepted standards also allows us to use off-the-shelf applications and tools that support them.
However, standards have their limitations. They are generic and tend to cover all possible cases. They are also often incomplete. In order to overcome this limitation, standards need to be profiled (i.e., customized) to the needs of a given community. As depicted on the Figure above, such a profile consists of both:
- Selecting a relevant subset of the standard to be profiled,
- Enriching this standard with components required by a specific community, which are often also borrowed piecemeal from other standards. (More on profiling can be found here)
Although perfectly legitimate, this type of customization breaks interoperability with the source standards and requires the development (and maintenance) of ad hoc solutions to support them.
This usually implies additional costs and increased risks.
We’ll explain in a next post how we handle these issues. In the meantime we would be glad to hear how others have tackled this problem.