On Aug 29, 2011, at 7:36 PM, Gerrit wrote:

>>> Basically what I'm trying to do is to remove triples that were
>>> previously inferred (based on a CONSTRUCT SPIN rule), when the triple
>>> which caused the inferment does not exist anymore. This should be
>>> quite a common use case, no?
>> 
>> Why are those triples created in the first place? Why don't you just prevent 
>> their creation by modifying the SPIN rules - make the WHERE clause of that 
>> rule more specific, it's open source after all.
> 
> The triples are inferred (using SPIN rules) when new triples are
> inserted by devices connected to a triple store. As an example,
> consider a device inserting a sc:connectedTo relationship when it is
> connected to another device. When the asserted triples are later
> removed by these devices (e.g. a device removing its sc:connectedTo
> relationship when the other device disconnects), the triples that were
> inferred based on these triples should be removed as well.

Ah yes, this is a tricky use case. TBC has some approximation of such behavior 
called incremental inferencing in TopSPIN, but this has some assumptions that 
all rules related to a resource are attached to its class. In the case of the 
OWL RL triples this would be too difficult because it would re-run all rules 
each time. What TopSPIN incremental is doing, is to run the relevant rules once 
with and once without the deleted triple, and then also removes inferences that 
depended on the deleted triple. This assumes that there is a change set 
describing the deleted triple and also uses some classes (BufferingGraphs) that 
are not part of the standard API. So I have no additional insights at this 
stage.

Holger

-- 
You received this message because you are subscribed to the Google
Group "TopBraid Suite Users", the topics of which include TopBraid Composer,
TopBraid Live, TopBraid Ensemble, SPARQLMotion and SPIN.
To post to this group, send email to
[email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/topbraid-users?hl=en

Reply via email to