aboutsummaryrefslogtreecommitdiff
path: root/src/main
Commit message (Collapse)AuthorAgeFilesLines
* Prepare code for double-blind submissionv0.1.0Federico Igne2021-04-153-14/+0
|
* Fix handling of class expression OWLObjectHasSelfFederico Igne2021-04-092-46/+54
| | | | It was not properly handled in the normalization process.
* Streamline RSA approximationFederico Igne2021-04-094-162/+132
|
* Add approximation to RSAFederico Igne2021-04-071-107/+159
| | | | | | | | Note that while the code is complete for approximating an input OWL2 ontology to RSA, the final steps are not "connected" yet, and the approximation won't fire automatically. We still need to find a way to include the approximation in the workflow in an efficient and transparent way.
* Add shifting for disjunction in the rhs of an axiomFederico Igne2021-04-051-2/+56
|
* Fix generation of fresh classes during normalizationFederico Igne2021-04-051-1/+4
|
* Fix minor issues with revised filtering programFederico Igne2021-04-022-80/+97
|
* Apply minor tweaksFederico Igne2021-04-022-11/+14
|
* Add revised implementation of filtering programFederico Igne2021-02-044-6/+414
|
* Add versioning system for different versions of codeFederico Igne2021-02-044-12/+66
| | | | | Later on this will allow us to select the algorithm from the command line and compare performance easily.
* Add ontology normalizerFederico Igne2021-02-043-2/+541
| | | | | This also allows to define ontology approximations to RSA in a simple way.
* Fix facts/rules import into RDFoxFederico Igne2021-01-301-8/+22
|
* Delay import of data files (#7)Federico Igne2021-01-306-194/+256
| | | | | This should partially solve the issue with data import through OWLAPI being too slow.
* Introduce a better system to handle command line inputFederico Igne2021-01-301-58/+142
|
* Simplify top axiomatization as shown in #7Federico Igne2021-01-261-43/+84
|
* Add OWLDataHasValue class expression supportFederico Igne2021-01-231-1/+8
|
* Add OWLSymmetricObjectPropertyAxiom and OWLSubDataPropertyOfAxiom supportFederico Igne2021-01-231-5/+17
|
* Add OWLDisjointClassesAxiom supportFederico Igne2021-01-231-1/+7
|
* Add OWL{Data,Object}MinCardinality and OWLHasValue expression supportFederico Igne2021-01-232-5/+81
|
* Fix support for a few axiom types.Federico Igne2021-01-212-36/+22
| | | | | | | Namely: + OWLEquivalentClassesAxiom + OWLFunctionalObjectPropertyAxiom + OWLInverseFunctionalObjectPropertyAxiom
* Add external script to execute benchmarks and gather results.Federico Igne2021-01-053-9/+11
|
* Slightly rework main executionFederico Igne2021-01-051-46/+39
|
* Attempt to force a better ordering of body atoms in filtering rulesFederico Igne2021-01-051-9/+9
| | | | | | | | | This is based on an email from Yavor, suggesting a couple of improvements on the code. In particular this refers to the fact that, when RDFox lacks any information it assumes user ordering in the processing order of a rule body.
* Introduce code for transactions (commented out)Federico Igne2020-12-171-4/+17
| | | | | | | | | At this time transaction are not working well with the program, and importing everything in one transaction causes the query answering program to take too long. Hopefully we will be able to fix the bug and enable transactional execution of the major steps in the algorithm.
* Add additional debug information to query executionFederico Igne2020-12-171-10/+22
|
* Port code to RDFox v4.1.0Federico Igne2020-12-174-34/+61
|
* Remove old logic program conversion codeFederico Igne2020-12-113-370/+0
|
* Fix answer retrieval for boolean CQsFederico Igne2020-12-111-1/+1
|
* Merge branch 'logger' into masterFederico Igne2020-12-095-143/+298
|\
| * Keep track of answer multiplicityFederico Igne2020-12-094-17/+30
| |
| * Merge branch 'master' into loggerFederico Igne2020-12-084-11/+24
| |\
| * | Remove (forced) materializationFederico Igne2020-12-081-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | This is a useless call since RDFox automatically materialize it's datasets when committing a transactions. Since loading rules/facts uses an internal transaction internally, these materialization calls do nothing. It will be still usefull when operating inside transactions.
| * | Add diagnostics for (un)filtered answer ratioFederico Igne2020-12-073-5/+38
| | |
| * | Fix typo in ontology diagnosticsFederico Igne2020-12-071-2/+2
| | |
| * | Merge branch 'master' into loggerFederico Igne2020-12-072-2/+2
| |\ \
| * | | Add additional diagnosticsFederico Igne2020-12-072-1/+11
| | | |
| * | | Add first implementation of a loggerFederico Igne2020-12-074-129/+230
| | | | | | | | | | | | | | | | | | | | This is needed mostly for debugging and benchmarking purposes. We tried to make it as "out of the way" as possible.
* | | | Change how inverse roles are handle on the logic programming sideFederico Igne2020-12-081-1/+2
| |_|/ |/| | | | | | | | | | | | | | | | | | | | Seems like an easy way to convert inverse roles into atoms is to write `R(y,x)` instead of `R^-(x,y)`. This is still something we need to test and some additional rules that use inverse roles in the naive way are still generated.
* | | Fix bug for OWLObjectSubPropertyOfAxiomFederico Igne2020-12-081-5/+4
| | | | | | | | | | | | | | | A version of the rule with an empty suffix was missing. Note that this is not completely clear from the paper.
* | | Include literals in instances of `Named` internal predicateFederico Igne2020-12-082-5/+19
| | |
* | | Fix bug for data property translationFederico Igne2020-12-071-1/+1
| |/ |/|
* | Change ConjunctiveQuery constructors interfaceFederico Igne2020-12-073-3/+3
|/ | | | | One of the object constructor was returning an Option. It was weird to have a constructor that would not always return an object.
* Update answers pretty printerFederico Igne2020-12-071-2/+7
|
* Change answer pritty printer to use CSV-like syntaxFederico Igne2020-12-061-4/+8
|
* Add support for data property assertion axioms (OWLDataPropertyAssertionAxiom)Federico Igne2020-12-062-8/+95
|
* Reintroduce previous implementation of `cycle`Federico Igne2020-12-061-29/+32
| | | | | At the time of writing, the naïve implementation of `cycle` produces too many rules, which are not efficiently handled by RDFox.
* Add method to retrieve diagnostics from RDFoxFederico Igne2020-12-061-0/+15
|
* Fix axiom skolemizationFederico Igne2020-12-041-3/+3
| | | | | | Skolemization should be uniquely bound to the axiom being skolemized. When an axiom is a shortcut for a set of axioms, the skolemization strategy should be updated with the new axiom(s).
* Simplify duplication of skolemization strategyFederico Igne2020-12-041-6/+10
|
* Rework skolemization strategiesFederico Igne2020-12-047-161/+125
| | | | | | In particular `ConstantRSA` has been removed, since it was *not* a skolemization strategy. The case for extra atoms generation previously handled by `ConstantRSA` is not dealt with inside the RSA check.
generated by cgit v1.2.3 (git 2.25.1) at 2025-07-13 19:42:51 +0000