Extends the Graph interface to give additional means to query an inferred
graph. Many entailments from the raw data are made to appear as if they
are extract triples in the inferred graph and so appear through the
normal Graph.find interface.
However, here are two extensions required. Firstly, the ability to
ask about global properties of the whole graph (e.g. consistency). Secondly,
the ability to temporarily construct expressions (encoded in RDF) which
form more complex queries.
Fields inherited from interface org.apache.jena.graph.Graph
Replace the underlying data graph for this inference graph and start any
inferences over again. This is primarily using in setting up ontology imports
processing to allow an imports multiunion graph to be inserted between the
inference graph and the raw data, before processing.
data - the new raw data graph
Cause the inference graph to reconsult the underlying graph to take
into account changes. Normally changes are made through the InfGraph's add and
remove calls are will be handled appropriately. However, in some cases changes
are made "behind the InfGraph's back" and this forces a full reconsult of
the changed data.
Perform any initial processing and caching. This call is optional. Most
engines either have negligible set up work or will perform an implicit
"prepare" if necessary. The call is provided for those occasions where
substantial preparation work is possible (e.g. running a forward chaining
rule system) and where an application might wish greater control over when
this preparation is done.
Reset any internal caches. Some systems, such as the tabled backchainer,
retain information after each query. A reset will wipe this information preventing
unbounded memory use at the expense of more expensive future queries. A reset
does not cause the raw data to be reconsulted and so is less expensive than a rebind.
Test a global boolean property of the graph. This might included
properties like consistency, OWLSyntacticValidity etc.
It remains to be seen what level of generality is needed here. We could
replace this by a small number of specific tests for common concepts.
property - the URI of the property to be tested
a Node giving the value of the global property, this may
be a boolean literal, some other literal value (e.g. a size).
An extension of the Graph.find interface which allows the caller to
encode complex expressions in RDF and then refer to those expressions
within the query triple. For example, one might encode a class expression
and then ask if there are any instances of this class expression in the
subject - the subject Node of the query triple, may be a Node in
the graph or a node in the parameter micro-graph or null
property - the property to be retrieved or null
object - the object Node of the query triple, may be a Node in
the graph or a node in the parameter micro-graph.
param - a small graph encoding an expression which the subject and/or
object nodes refer.
Returns a derivations graph. The rule reasoners typically create a
graph containing those triples added to the base graph due to rule firings.
In some applications it can useful to be able to access those deductions
directly, without seeing the raw data which triggered them. In particular,
this allows the forward rules to be used as if they were rewrite transformation
the deductions graph, if relevant for this class of inference
engine or null if not.