ctakes-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "John Green" <john.travis.gr...@gmail.com>
Subject Re: groovy script(s) for running equivalent of AggregatePlaintextUMLSProcessor.xml
Date Wed, 02 Apr 2014 15:24:09 GMT
Finally got around to looking at these: work great on ubuntu 13.1




JG

—
Sent from Mailbox for iPhone

On Tue, Dec 24, 2013 at 3:48 PM, Masanz, James J. <Masanz.James@mayo.edu>
wrote:

> I checked in 2 groovy scripts under
> ctakes-core\scripts\groovy\
> I checked into trunk. Will gladly move or remove if needed.
> They do not do automatic downloads of ctakes, instead they rely on user to already have
downloaded a convenience binary and the separately downloadable resources, and to have extracted
both of those.
> They run the same as what's defined in AggregatePlaintextUMLSProcessor.xml except I haven't
included the semantic role labeler or the dependency parser yet (just for lack of time today).
 Right now the script uses XML descriptors for some parts and uimaFIT without using XML descriptor
files for other parts of the pipeline.
> I have a few other TODOs but the scripts are ready for others to look at.
> Feedback welcome and encouraged!
> To run:
> Install groovy
> Have groovy on your path.
> Within run_cTAKES.groovy, you need to set the values of
> cTAKES_HOME
> EXTERNAL_RESOURCE_PATH
> Set  JAVA_OPTS env variable as follows
> -Dctakes.umlsuser=YOUR_USERID  -Dctakes.umlspw=YOUR_PWD
> Then to run, enter this, with inputDirName being wherever you have your input text files.
> groovy run_cTAKES.groovy inputDirName
> -- James
Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message