The problem was caused by two separate bugs in the rule inference. They
are now fixed in the current trunk.
Thanks for reporting it :-)
Best,
Peter
Am 10.07.2015 um 15:46 schrieb Peter Klügl:
Hi,
Am 10.07.2015 um 13:55 schrieb Andreas Weber:
Peter Klügl <pkluegl@...> writes:
... in the light of current events and since you use it integrated in
your software:
Did you use an analysis engine for creating the CAS or, in case you
created it using a type system, did you consider the type priotities?
Hi Peter,
thanks for your help.
We create our CAS and annotations programmatically using our own type
system, we don't use another UIMA annotator.
I played a little bit around with type priorities (MY_ANNO > W) but
that
had no impact. The main problem seems to be that we have the _same_
annotation type at one input position.
That shouldnt't be a problem at all ;-)
UIMA Ruta brings its own type priorities for some special internal
types. If the priorities are missing then there could be some problems
in sequential matching.
I'll take a look at it.
Best,
Peter
Regards,
Andreas
