To validate our hypothesis about a graph split, we need SPARQL endpoints exposing both the split graphs and the full graph so that we can compare query results and performances. We want to expose both the full graph and the split graphs, with the same data (based on the same Wikidata dump) so that comparisons are meaningful.
Actual work will be tracked as sub tasks.
AC:
- 3 servers are loaded with:
- full graph
- scholarly articles
- Full graph (excluding scholarly articles)
- SPARQL endpoints are exposed publicly so that our users can experiment