The SPARQL CONSTRUCT query to turn BF2.0 into Schema is quite slow. It is currently the slowest part of the conversion pipeline. I think there's room to optimize, for example analyzing internal duplication of result rows due to cross products in the query.
The SPARQL CONSTRUCT query to turn BF2.0 into Schema is quite slow. It is currently the slowest part of the conversion pipeline. I think there's room to optimize, for example analyzing internal duplication of result rows due to cross products in the query.