Identified challenges while working out the process
- Niek van Ulzen
- Ana Konrad
status: in development
While working out and documenting the process for developing domain-specific metadata schemas we identified several challenges concerning either contributing to working out the process or to schema development. Below we listed these challenges. These will be addressed while working on the process and (potential) solutions/actions will be described here as well.
No# | Challenge |
|---|---|
01 | We as a domain already thought about metadata models for years and we already have an information model. Why not reuse that instead of mapping that again to a new model? |
02 | Developing a (domain-specific) metadata schema is not a linear process, but much more an iterative process |
03 | How to guarantee long-term governance of metadata schemas? |
04 | It is not clear to us what is in the core/health ring and what should be in a petal. Maybe all of our domain wishes are coverered by Health-DCAT-AP |
05 | We don’t know if, as a domain or working group, we have the capacity and/or expertise to contribute to documenting the process or start developing a domain-specific metadata schema |
06 | An end user like a reseacher may experience more added value on the data modeling level (doing ‘deep search’) than on the metadata-for-catalogue level |
07 | When we talk about (domain-specific) metadata, it seems we are talking about different things |
08 | We don’t know exactly which (sub)domain we are representing and if we have the mandate to speak on behalf of that (sub)domain and if we can make decisions on our own |
09 | We don’t think there will be lots of enthusiasm about or added value of only writing down the process steps ‘from theory’ before actually starting developing and writing down the process from experiences/practice |
10 | What happens if a certain metadata property overlaps for different domains, but each domain uses a different label/ontology term to describe it? Or worse: the same term with conflicting definitions. For instance, see the literature for “false-agreements” |
11 | How should we take into consideration what searches and metadata properties are important for an end user? |
12 | How do we relate to international domain-specific developments that go faster or slower than us? In some cases, domain standards, schemas, or models are still evolving, which can pose an even greater challenge in reaching consensus. |
13 | It might be that for some domain-specific classes and properties we cannot reuse already existing ontologies. Do we need to create our own ontology, how to do that, and how to deal with our own namespaces? |