I have successfully (I think) managed to build/install/configure cTAKES 3.1.2 +ytex using the provided directions, Maven and Ant scripts, and MetamorphoSys scripts. This includes populating a MySQL DB with the UMLS data (thanks to MetamorphoSys) and using that to populate the ytex schema in that DB (thanks to the ytex Ant script). The ytex schema now includes a v_snomed_fword_lookup table containing 567,855 records, for example. I’ve edited the runctakesCVD.bat file to add my UMLS user name and pwd just as I did with cTAKES 3.1.1. However, when I run that bat file, tell the GUI to load the AE at apache-ctakes-3.1.2-SNAPSHOT\desc\ctakes-ytex-uima\desc\analysis_engine\AggregatePlaintextUMLSProcessor.xml, and try to process various text files, the resulting annotations in org.apache.ctakes.typesystem.type.textsem.IdentifiedAnnotation are lacking all the medical annotations (such as DiseaseDisorderMention) I would expect (and see if I use cTAKES 3.1.1). There’s the occasional DateAnnotation, FractionAnnotation, etc, but that’s it. The only indication of possible trouble is a whole bunch of log entries like: “DEBUG [FirstTokenPermutationImpl] Window size of 8 exceeds the max permutation level of 7.” I’m not sure if that explains the problem I’m having. If I try to instead use the AE at apache-ctakes-3.1.2-SNAPSHOT\desc\ctakes-clinical-pipeline\desc\analysis_engine\AggregatePlaintextUMLSProcessor.xml and run a document, I get the exception “org.apache.uima.analysis_engine.AnalysisEngineProcessException: More detailed information is in the log file.” But the log file does not show any problems even though I’ve jacked up the logging level. There are earlier entries like: “DEBUG [DataBinder] DataBinder requires binding of required fields []” … which make me nervous but leave me in the dark. Any suggestions? Thanks.
