Hello,
we do not have a PEAR distribution, because we cannot distribute our models.
There is an ant script which can build a PEAR from the sample
descriptors and the
models from sourceforge, the script downloads the models.
In my UIMA projects I use maven to depend on the OpenNLP jars and keep
the descriptors out of the jars. My models are usually hosted on a web
server
and I then just reference them by URL in the UIMA descriptors.
Jörn
On 10/25/2012 02:12 PM, Yasen Kiprov wrote:
Hello,
I'm trying to setup a UIMA pipeline with the opennlp tools and I have some
questions. There are AE descriptors in the opennlp-uima project but they are
not available in the maven artifact (i.e. here:
http://repo1.maven.org/maven2/org/apache/opennlp/opennlp-uima/1.5.2-incubating/).
Is there a PEAR-like distribution of the opennlp tools or should I use the
opennlp-dist? Or should I just copy the descriptors in my project and depend on
opennlp-uima?
What is the best way to depend on the opennlp when building UIMA projects in
maven?
Thank you,
Yasen
