Task: Test performance of doing conversion via SPARQL Anything

[mwigham]

SPARQL Anything offers us an alternative way of doing the mapping of our metadata from json to RDF. The question is how good the performance would be, especially given our large number of metadata fields, which would require multiple SPARQL Anything queries to keep things readable and maintainable.

We need to choose an appropriate moment to investigate this, as it would require a large amount of refactoring in the code if we do indeed decide to switch to SPARQL Anything.

Tasks:

• [ ] Define part of our mapping in SPARQL Anything (See showcase-tate)
• [ ] Test the performance
• [ ] Estimate what the performance would be for the complete mapping
• [ ] Compare to the results of looking at RML
• [ ] Decide whether to make the switch to SPARQL Anything (this would be done in a new series of issues)

[wmelder]

and maybe refresh the RML components and compare to that?

[mwigham]

Good idea, I'll add it to the task list

[mwigham]

on second thoughts, I'll make another issue