This content has been marked as final. Show 1 reply
We are actually looking into a general mechanism that allows users to plug in their own queries/logic during inference. Thanks very much for bringing SPIN up. This general mechanism is very likely going to cover CONSTRUCT queries.
To extend the existing inference engine using the existing Jena Adapter release, one possible way is as follows.
1) Assume you have an ontology (model) A.
2) Create an empty model B.
3) run performInference on both A and B using OWLPrime.
4) run SPARQL CONSTRUCT queries against A, B and inferred data
5) store the query results (in the form of Jena models) back into model B.
6) If the size of model B does not change, then stop. Otherwise, repeat 3)
Note that model B is created to separate your original asserted data from inferred data.
If you don't need such a separation, then don't create it.