ctakes-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From chen...@apache.org
Subject svn commit: r1465043 [1/2] - in /ctakes/trunk: ./ ctakes-assertion/desc/analysis_engine/ ctakes-clinical-pipeline/desc/analysis_engine/ ctakes-clinical-pipeline/desc/analysis_engine/assertion_training_steps/ ctakes-clinical-pipeline/resources/launch/ c...
Date Fri, 05 Apr 2013 16:53:09 GMT
Author: chenpei
Date: Fri Apr  5 16:53:06 2013
New Revision: 1465043

URL: http://svn.apache.org/r1465043
Log:
CTAKES-96 - Update Dependency Parser to use new ClearNLP.

Added:
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/mayo-en-dep-1.3.0.jar   (with props)
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/dictionary-1.3.1.jar   (with props)
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/pred/
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/pred/mayo-en-pred-1.3.0.jar   (with props)
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/role/
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/role/mayo-en-role-1.3.0.jar   (with props)
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/mayo-en-srl-1.3.0.jar   (with props)
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserPlaintextAggregate.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserPlaintextAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedAggregate.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedInfPosAggregate.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedInfPosAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLPlaintextAggregate.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLPlaintextAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedAggregate.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedInfPosAggregate.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedInfPosAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSemanticRoleLabelerAE.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSemanticRoleLabelerAE.xml
    ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearNLPTestCPE.xml
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearParserTestCPE.xml
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearNLPDependencyParserAE.java
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearParserDependencyParserAE.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearNLPSemanticRoleLabelerAE.java
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearParserSemanticRoleLabelerAE.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/util/TestClearNLPAnalysisEngines.java
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/util/TestClearParserAnalysisEngines.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/util/WriteClearNLPDescriptors.java
      - copied, changed from r1464565, ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/util/WriteClearParserDescriptors.java
Removed:
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserPlaintextAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedInfPosAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLPlaintextAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedInfPosAggregate.xml
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSemanticRoleLabelerAE.xml
    ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearParserTestCPE.xml
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearParserDependencyParserAE.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearParserSemanticRoleLabelerAE.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearParserUtil.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/LemAssigner.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/PosAssigner.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/util/TestClearParserAnalysisEngines.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/util/WriteClearParserDescriptors.java
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/util/DependencyRegex.java
Modified:
    ctakes/trunk/ctakes-assertion/desc/analysis_engine/AttributeDiscoveryPreprocessor.xml
    ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaProcessor.xml
    ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaUMLSProcessor.xml
    ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextProcessor.xml
    ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextUMLSProcessor.xml
    ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/assertion_training_steps/AggregatePlaintextUMLSProcessor__no_dictionary_no_assertion.xml
    ctakes/trunk/ctakes-clinical-pipeline/resources/launch/UIMA_CVD--clinical_documents_pipeline.launch
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/README
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/README
    ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/README
    ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/README.txt
    ctakes/trunk/ctakes-dependency-parser/pom.xml
    ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/util/ClearDependencyUtility.java
    ctakes/trunk/ctakes-regression-test/desc/analysis_engine/RegressionPipelineAggregateTest.xml
    ctakes/trunk/ctakes-relation-extractor/desc/analysis_engine/RelationExtractorAggregate.xml
    ctakes/trunk/ctakes-relation-extractor/desc/analysis_engine/RelationExtractorPreprocessor.xml
    ctakes/trunk/ctakes-temporal/src/main/java/org/apache/ctakes/temporal/eval/Evaluation_ImplBase.java
    ctakes/trunk/ctakes-utils/src/main/java/org/apache/ctakes/utils/build/BuildPear.dependency parser.properties
    ctakes/trunk/pom.xml

Modified: ctakes/trunk/ctakes-assertion/desc/analysis_engine/AttributeDiscoveryPreprocessor.xml
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-assertion/desc/analysis_engine/AttributeDiscoveryPreprocessor.xml?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-assertion/desc/analysis_engine/AttributeDiscoveryPreprocessor.xml (original)
+++ ctakes/trunk/ctakes-assertion/desc/analysis_engine/AttributeDiscoveryPreprocessor.xml Fri Apr  5 16:53:06 2013
@@ -4,7 +4,7 @@
   <primitive>false</primitive>
   <delegateAnalysisEngineSpecifiers>
     <delegateAnalysisEngine key="DependencyParser">
-      <import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml"/>
+      <import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml"/>
     </delegateAnalysisEngine>
     <delegateAnalysisEngine key="Chunker">
       <import location="../../../ctakes-chunker/desc/Chunker.xml"/>

Modified: ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaProcessor.xml
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaProcessor.xml?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaProcessor.xml (original)
+++ ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaProcessor.xml Fri Apr  5 16:53:06 2013
@@ -69,7 +69,7 @@
 <import location="../../../ctakes-assertion/desc/AssertionMiniPipelineAnalysisEngine.xml"/>
 </delegateAnalysisEngine>
 <delegateAnalysisEngine key="DependencyParser">
-<import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml"/>
+<import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml"/>
 </delegateAnalysisEngine>
 </delegateAnalysisEngineSpecifiers>
 <analysisEngineMetaData>

Modified: ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaUMLSProcessor.xml
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaUMLSProcessor.xml?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaUMLSProcessor.xml (original)
+++ ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregateCdaUMLSProcessor.xml Fri Apr  5 16:53:06 2013
@@ -69,7 +69,7 @@
 <import location="../../../ctakes-assertion/desc/AssertionMiniPipelineAnalysisEngine.xml"/>
 </delegateAnalysisEngine>
 <delegateAnalysisEngine key="DependencyParser">
-<import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml"/>
+<import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml"/>
 </delegateAnalysisEngine>
 </delegateAnalysisEngineSpecifiers>
 <analysisEngineMetaData>

Modified: ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextProcessor.xml
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextProcessor.xml?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextProcessor.xml (original)
+++ ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextProcessor.xml Fri Apr  5 16:53:06 2013
@@ -69,7 +69,10 @@
 <import location="../../../ctakes-assertion/desc/AssertionMiniPipelineAnalysisEngine.xml"/>
 </delegateAnalysisEngine>
 <delegateAnalysisEngine key="DependencyParser">
-<import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml"/>
+<import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml"/>
+</delegateAnalysisEngine>
+<delegateAnalysisEngine key="SemanticRoleLabeler">
+<import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearNLPSemanticRoleLabelerAE.xml"/>
 </delegateAnalysisEngine>
 </delegateAnalysisEngineSpecifiers>
 <analysisEngineMetaData>
@@ -121,6 +124,8 @@ Includes adjusting NP chunks to include 
 <node>LookupWindowAnnotator</node>
 <node>DictionaryLookupAnnotator</node>
 <node>DependencyParser</node>
+<node>SemanticRoleLabeler</node>
+
 <node>AssertionAnnotator</node>
 <!-- 
 	<node>StatusAnnotator</node>

Modified: ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextUMLSProcessor.xml
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextUMLSProcessor.xml?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextUMLSProcessor.xml (original)
+++ ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/AggregatePlaintextUMLSProcessor.xml Fri Apr  5 16:53:06 2013
@@ -69,8 +69,11 @@
       <import location="../../../ctakes-assertion/desc/AssertionMiniPipelineAnalysisEngine.xml"/>
     </delegateAnalysisEngine>
     <delegateAnalysisEngine key="DependencyParser">
-      <import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml"/>
+      <import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml"/>
     </delegateAnalysisEngine>
+<delegateAnalysisEngine key="SemanticRoleLabeler">
+<import location="../../../ctakes-dependency-parser/desc/analysis_engine/ClearNLPSemanticRoleLabelerAE.xml"/>
+</delegateAnalysisEngine>    
   </delegateAnalysisEngineSpecifiers>
   <analysisEngineMetaData>
     <name>AggregatePlaintextUMLSProcessor</name>
@@ -122,6 +125,7 @@ and requires an UMLS license.  Please up
         <node>LookupWindowAnnotator</node>
         <node>DictionaryLookupAnnotatorDB</node>
         <node>DependencyParser</node>
+		<node>SemanticRoleLabeler</node>        
         <node>AssertionAnnotator</node>
         <!-- 
         	<node>StatusAnnotator</node>

Modified: ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/assertion_training_steps/AggregatePlaintextUMLSProcessor__no_dictionary_no_assertion.xml
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/assertion_training_steps/AggregatePlaintextUMLSProcessor__no_dictionary_no_assertion.xml?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/assertion_training_steps/AggregatePlaintextUMLSProcessor__no_dictionary_no_assertion.xml (original)
+++ ctakes/trunk/ctakes-clinical-pipeline/desc/analysis_engine/assertion_training_steps/AggregatePlaintextUMLSProcessor__no_dictionary_no_assertion.xml Fri Apr  5 16:53:06 2013
@@ -80,7 +80,7 @@
     </delegateAnalysisEngine>
     -->
     <delegateAnalysisEngine key="DependencyParser">
-      <import location="../../../../ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml"/>
+      <import location="../../../../ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml"/>
     </delegateAnalysisEngine>
   </delegateAnalysisEngineSpecifiers>
   <analysisEngineMetaData>

Modified: ctakes/trunk/ctakes-clinical-pipeline/resources/launch/UIMA_CVD--clinical_documents_pipeline.launch
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-clinical-pipeline/resources/launch/UIMA_CVD--clinical_documents_pipeline.launch?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-clinical-pipeline/resources/launch/UIMA_CVD--clinical_documents_pipeline.launch (original)
+++ ctakes/trunk/ctakes-clinical-pipeline/resources/launch/UIMA_CVD--clinical_documents_pipeline.launch Fri Apr  5 16:53:06 2013
@@ -13,5 +13,5 @@
 <stringAttribute key="org.eclipse.jdt.launching.MAIN_TYPE" value="org.apache.uima.tools.annot_view.Gladis"/>
 <stringAttribute key="org.eclipse.jdt.launching.PROJECT_ATTR" value="ctakes-clinical-pipeline"/>
 <stringAttribute key="org.eclipse.jdt.launching.SOURCE_PATH_PROVIDER" value="org.eclipse.m2e.launchconfig.sourcepathProvider"/>
-<stringAttribute key="org.eclipse.jdt.launching.VM_ARGUMENTS" value="&quot;-Djava.util.logging.config.file=${env_var:UIMA_HOME}/Logger.properties&quot;  -Xms500M -Xmx1g"/>
+<stringAttribute key="org.eclipse.jdt.launching.VM_ARGUMENTS" value="&quot;-Djava.util.logging.config.file=${env_var:UIMA_HOME}/Logger.properties&quot;  -XX:+UseConcMarkSweepGC -Xms500M -Xmx1500M"/>
 </launchConfiguration>

Modified: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/README
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/README?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/README (original)
+++ ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/README Fri Apr  5 16:53:06 2013
@@ -1,37 +1,14 @@
-############################
-conll-2009-dev-shift-pop.jar
-############################
-The model conll-2009-dev-shift-pop.jar was provided by Jinho Choi and is trained on the 
-development set of the CoNLL 2009 English data.  More information about this data can 
-be found here:
-
-http://ufal.mff.cuni.cz/conll2009-st/train-dev-data.html
-
-The ConLL 2009 shared task is described in the following paper. 
-http://aclweb.org/anthology-new/W/W09/W09-1201.pdf
-
-This model uses the "shift pop" algorithm.  When using this model you should specify the 
-value AbstractDepParser.ALG_SHIFT_POP for the configuration parameter 
-org.cleartk.syntax.dependency.clear.ClearParser.parserAlgorithmName as annotated in 
-ClearParser.  Also, your default memory setting for your JVM may not suffice to load this 
-model.  It can load with the following argument "-Xmx1g". 
-
-On 29 Sep 2011, the model was modified by adding a "1" before the "17" near the end of
-the "lexica" file, to accomodate a change in model format for the 0.4.0-SNAPSHOT release.
-
 #################
 additional models
 #################
 
 There is an additional model available from the ClearTK downloads page:
 
-http://code.google.com/p/cleartk/downloads/list
+https://code.google.com/p/clearnlp/
 
-and is called conll-2009-training-dev-shift-pop.jar.  This model is built from
+and is called mayo-en-dep-1.3.0.jar.  This model is built from
 both the training data and development data from the CoNLL 2009 shared task (see
-links above).  This model was provided by Jinho Choi as a file named 
-conll-trndev-sp.mod.3.  This model is very large and expands considerably in memory. 
-You will need at least 8GB (gigabytes) to load it and run it.     
+links above).  This model was provided by Jinho Choi.
 
 #####
 Notes

Added: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/mayo-en-dep-1.3.0.jar
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/mayo-en-dep-1.3.0.jar?rev=1465043&view=auto
==============================================================================
Binary file - no diff available.

Propchange: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/dependency/mayo-en-dep-1.3.0.jar
------------------------------------------------------------------------------
    svn:mime-type = application/octet-stream

Modified: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/README
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/README?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/README (original)
+++ ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/README Fri Apr  5 16:53:06 2013
@@ -1,9 +1,2 @@
-The file wordnet-3.0-lemma-data.jar is a data file that can be used by LemmaAnnotator.  
-It was built using WordNet version 3.0 data and is therefore subject to any 
-restrictions imposed by the WordNet license which appears to be very similar
-to the BSD license (both WordNet software and database.)
-
-http://wordnet.princeton.edu/wordnet/license/
-
-The file wordnet-3.0-lemma-data.jar was obtained directly from the author of
-clear.morph.MorphEnAnalyzer, Jinho Choi, and was called en_dict-1.0.jar. 
\ No newline at end of file
+The file dictionary-1.3.0.jar was obtained directly from the author of
+clearn NLP, Jinho Choi.
\ No newline at end of file

Added: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/dictionary-1.3.1.jar
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/dictionary-1.3.1.jar?rev=1465043&view=auto
==============================================================================
Binary file - no diff available.

Propchange: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/lemmatizer/dictionary-1.3.1.jar
------------------------------------------------------------------------------
    svn:mime-type = application/octet-stream

Added: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/pred/mayo-en-pred-1.3.0.jar
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/pred/mayo-en-pred-1.3.0.jar?rev=1465043&view=auto
==============================================================================
Binary file - no diff available.

Propchange: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/pred/mayo-en-pred-1.3.0.jar
------------------------------------------------------------------------------
    svn:mime-type = application/octet-stream

Added: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/role/mayo-en-role-1.3.0.jar
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/role/mayo-en-role-1.3.0.jar?rev=1465043&view=auto
==============================================================================
Binary file - no diff available.

Propchange: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/role/mayo-en-role-1.3.0.jar
------------------------------------------------------------------------------
    svn:mime-type = application/octet-stream

Modified: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/README
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/README?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/README (original)
+++ ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/README Fri Apr  5 16:53:06 2013
@@ -1,7 +1,7 @@
 ############################
-en_srl_ontonotes.jar
+mayo-en-srl-1.3.0.jar, ../pred/mayo-en,pred-1.3.0.jar, ../role/mayo-en-role-1.3.0.jar
 ############################
-The model en_srl_ontonotes.jar was provided by Jinho Choi and is trained on the 
+The model mayo-en-srl-1.3.0.jar was provided by Jinho Choi and is trained on the 
 development set of the OntoNotes corpus version 4.0.  More information about this data can 
 be found here:
 http://www.ldc.upenn.edu/Catalog/docs/LDC2011T03/OntoNotes-Release-4.0.pdf
@@ -21,8 +21,8 @@ for the token ")" should be ")".)  This 
 may use tags such as "COLON" or "RRB".  Your part-of-speech tags should be modified to be consistent with
 the tagging scheme used here.
 
-- The models also depend on a dependency tree produced by the ClearParser dependency parser.  This should be
-run prior to running ClearParser semantic role labeler.
+- The models also depend on a dependency tree produced by the ClearNLP dependency parser.  This should be
+run prior to running ClearNLP semantic role labeler.
 
 - The dependency labeling scheme produced by the models provided here is separate/different from that of
 the Malt parser models used by ClearTK's wrapper of the Malt parser.  You should not assume that 

Added: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/mayo-en-srl-1.3.0.jar
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/mayo-en-srl-1.3.0.jar?rev=1465043&view=auto
==============================================================================
Binary file - no diff available.

Propchange: ctakes/trunk/ctakes-dependency-parser-res/src/main/resources/org/apache/ctakes/dependency/parser/models/srl/mayo-en-srl-1.3.0.jar
------------------------------------------------------------------------------
    svn:mime-type = application/octet-stream

Copied: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserAE.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserAE.xml Fri Apr  5 16:53:06 2013
@@ -22,30 +22,30 @@
 <analysisEngineDescription xmlns="http://uima.apache.org/resourceSpecifier">
     <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
     <primitive>true</primitive>
-    <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserDependencyParserAE</annotatorImplementationName>
+    <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPDependencyParserAE</annotatorImplementationName>
     <analysisEngineMetaData>
-        <name>ClearParserDependencyParserAE</name>
+        <name>ClearNLPDependencyParserAE</name>
         <description>Descriptor automatically generated by uimaFIT</description>
         <version>unknown</version>
         <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
         <configurationParameters>
             <configurationParameter>
                 <name>ParserModelFileName</name>
-                <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearParserUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
+                <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearNLPUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
                 <type>String</type>
                 <multiValued>false</multiValued>
                 <mandatory>false</mandatory>
             </configurationParameter>
             <configurationParameter>
                 <name>ParserAlgorithmName</name>
-                <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearParserUtil.  If in doubt, do not change from the default value.</description>
+                <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearNLPUtil.  If in doubt, do not change from the default value.</description>
                 <type>String</type>
                 <multiValued>false</multiValued>
                 <mandatory>true</mandatory>
             </configurationParameter>
             <configurationParameter>
                 <name>UseLemmatizer</name>
-                <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                 <type>Boolean</type>
                 <multiValued>false</multiValued>
                 <mandatory>false</mandatory>

Copied: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserPlaintextAggregate.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserPlaintextAggregate.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserPlaintextAggregate.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserPlaintextAggregate.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserPlaintextAggregate.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserPlaintextAggregate.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserPlaintextAggregate.xml Fri Apr  5 16:53:06 2013
@@ -23,34 +23,34 @@
     <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
     <primitive>false</primitive>
     <delegateAnalysisEngineSpecifiers>
-        <delegateAnalysisEngine key="ClearParserDependencyParserAE">
+        <delegateAnalysisEngine key="ClearNLPDependencyParserAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserDependencyParserAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPDependencyParserAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserDependencyParserAE</name>
+                    <name>ClearNLPDependencyParserAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearParserUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
+                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearNLPUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>ParserAlgorithmName</name>
-                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearParserUtil.  If in doubt, do not change from the default value.</description>
+                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearNLPUtil.  If in doubt, do not change from the default value.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>true</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -687,7 +687,7 @@
                 <node>LVG Annotator</node>
                 <node>POSTagger</node>
                 <node>Chunker</node>
-                <node>ClearParserDependencyParserAE</node>
+                <node>ClearNLPDependencyParserAE</node>
             </fixedFlow>
         </flowConstraints>
         <capabilities/>

Copied: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedAggregate.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedAggregate.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedAggregate.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedAggregate.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedAggregate.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedAggregate.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedAggregate.xml Fri Apr  5 16:53:06 2013
@@ -23,34 +23,34 @@
     <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
     <primitive>false</primitive>
     <delegateAnalysisEngineSpecifiers>
-        <delegateAnalysisEngine key="ClearParserDependencyParserAE">
+        <delegateAnalysisEngine key="ClearNLPDependencyParserAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserDependencyParserAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPDependencyParserAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserDependencyParserAE</name>
+                    <name>ClearNLPDependencyParserAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearParserUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
+                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearNLPUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>ParserAlgorithmName</name>
-                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearParserUtil.  If in doubt, do not change from the default value.</description>
+                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearNLPUtil.  If in doubt, do not change from the default value.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>true</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -384,7 +384,7 @@
             <fixedFlow>
                 <node>SimpleSegmentAnnotator</node>
                 <node>LVG Annotator</node>
-                <node>ClearParserDependencyParserAE</node>
+                <node>ClearNLPDependencyParserAE</node>
             </fixedFlow>
         </flowConstraints>
         <capabilities/>

Copied: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedInfPosAggregate.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedInfPosAggregate.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedInfPosAggregate.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedInfPosAggregate.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedInfPosAggregate.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserDependencyParserTokenizedInfPosAggregate.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPDependencyParserTokenizedInfPosAggregate.xml Fri Apr  5 16:53:06 2013
@@ -23,34 +23,34 @@
     <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
     <primitive>false</primitive>
     <delegateAnalysisEngineSpecifiers>
-        <delegateAnalysisEngine key="ClearParserDependencyParserAE">
+        <delegateAnalysisEngine key="ClearNLPDependencyParserAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserDependencyParserAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPDependencyParserAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserDependencyParserAE</name>
+                    <name>ClearNLPDependencyParserAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearParserUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
+                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearNLPUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>ParserAlgorithmName</name>
-                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearParserUtil.  If in doubt, do not change from the default value.</description>
+                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearNLPUtil.  If in doubt, do not change from the default value.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>true</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -461,7 +461,7 @@
                 <node>SimpleSegmentAnnotator</node>
                 <node>LVG Annotator</node>
                 <node>POSTagger</node>
-                <node>ClearParserDependencyParserAE</node>
+                <node>ClearNLPDependencyParserAE</node>
             </fixedFlow>
         </flowConstraints>
         <capabilities/>

Copied: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLPlaintextAggregate.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLPlaintextAggregate.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLPlaintextAggregate.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLPlaintextAggregate.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLPlaintextAggregate.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLPlaintextAggregate.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLPlaintextAggregate.xml Fri Apr  5 16:53:06 2013
@@ -23,34 +23,34 @@
     <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
     <primitive>false</primitive>
     <delegateAnalysisEngineSpecifiers>
-        <delegateAnalysisEngine key="ClearParserDependencyParserAE">
+        <delegateAnalysisEngine key="ClearNLPDependencyParserAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserDependencyParserAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPDependencyParserAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserDependencyParserAE</name>
+                    <name>ClearNLPDependencyParserAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearParserUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
+                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearNLPUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>ParserAlgorithmName</name>
-                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearParserUtil.  If in doubt, do not change from the default value.</description>
+                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearNLPUtil.  If in doubt, do not change from the default value.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>true</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -388,27 +388,27 @@
                 <resourceManagerConfiguration/>
             </taeDescription>
         </delegateAnalysisEngine>
-        <delegateAnalysisEngine key="ClearParserSemanticRoleLabelerAE">
+        <delegateAnalysisEngine key="ClearNLPSemanticRoleLabelerAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserSemanticRoleLabelerAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPSemanticRoleLabelerAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserSemanticRoleLabelerAE</name>
+                    <name>ClearNLPSemanticRoleLabelerAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the semantic role labeler model required by the factory method provided by ClearParserUtil.</description>
+                            <description>This parameter provides the file name of the semantic role labeler model required by the factory method provided by ClearNLPUtil.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -652,8 +652,8 @@
                 <node>TokenizerAnnotatorPTB</node>
                 <node>LVG Annotator</node>
                 <node>POSTagger</node>
-                <node>ClearParserDependencyParserAE</node>
-                <node>ClearParserSemanticRoleLabelerAE</node>
+                <node>ClearNLPDependencyParserAE</node>
+                <node>ClearNLPSemanticRoleLabelerAE</node>
             </fixedFlow>
         </flowConstraints>
         <capabilities/>

Copied: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedAggregate.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedAggregate.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedAggregate.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedAggregate.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedAggregate.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedAggregate.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedAggregate.xml Fri Apr  5 16:53:06 2013
@@ -23,34 +23,34 @@
     <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
     <primitive>false</primitive>
     <delegateAnalysisEngineSpecifiers>
-        <delegateAnalysisEngine key="ClearParserDependencyParserAE">
+        <delegateAnalysisEngine key="ClearNLPDependencyParserAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserDependencyParserAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPDependencyParserAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserDependencyParserAE</name>
+                    <name>ClearNLPDependencyParserAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearParserUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
+                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearNLPUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>ParserAlgorithmName</name>
-                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearParserUtil.  If in doubt, do not change from the default value.</description>
+                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearNLPUtil.  If in doubt, do not change from the default value.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>true</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -334,27 +334,27 @@
                 </resourceManagerConfiguration>
             </taeDescription>
         </delegateAnalysisEngine>
-        <delegateAnalysisEngine key="ClearParserSemanticRoleLabelerAE">
+        <delegateAnalysisEngine key="ClearNLPSemanticRoleLabelerAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserSemanticRoleLabelerAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPSemanticRoleLabelerAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserSemanticRoleLabelerAE</name>
+                    <name>ClearNLPSemanticRoleLabelerAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the semantic role labeler model required by the factory method provided by ClearParserUtil.</description>
+                            <description>This parameter provides the file name of the semantic role labeler model required by the factory method provided by ClearNLPUtil.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -445,8 +445,8 @@
             <fixedFlow>
                 <node>SimpleSegmentAnnotator</node>
                 <node>LVG Annotator</node>
-                <node>ClearParserDependencyParserAE</node>
-                <node>ClearParserSemanticRoleLabelerAE</node>
+                <node>ClearNLPDependencyParserAE</node>
+                <node>ClearNLPSemanticRoleLabelerAE</node>
             </fixedFlow>
         </flowConstraints>
         <capabilities/>

Copied: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedInfPosAggregate.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedInfPosAggregate.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedInfPosAggregate.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedInfPosAggregate.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedInfPosAggregate.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSRLTokenizedInfPosAggregate.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSRLTokenizedInfPosAggregate.xml Fri Apr  5 16:53:06 2013
@@ -23,34 +23,34 @@
     <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
     <primitive>false</primitive>
     <delegateAnalysisEngineSpecifiers>
-        <delegateAnalysisEngine key="ClearParserDependencyParserAE">
+        <delegateAnalysisEngine key="ClearNLPDependencyParserAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserDependencyParserAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPDependencyParserAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserDependencyParserAE</name>
+                    <name>ClearNLPDependencyParserAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearParserUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
+                            <description>This parameter provides the file name of the dependency parser model required by the factory method provided by ClearNLPUtil.  If not specified, this analysis engine will use a default model from the resources directory</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>ParserAlgorithmName</name>
-                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearParserUtil.  If in doubt, do not change from the default value.</description>
+                            <description>This parameter provides the algorithm name used by the dependency parser that is required by the factory method provided by ClearNLPUtil.  If in doubt, do not change from the default value.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>true</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -334,27 +334,27 @@
                 </resourceManagerConfiguration>
             </taeDescription>
         </delegateAnalysisEngine>
-        <delegateAnalysisEngine key="ClearParserSemanticRoleLabelerAE">
+        <delegateAnalysisEngine key="ClearNLPSemanticRoleLabelerAE">
             <analysisEngineDescription>
                 <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
                 <primitive>true</primitive>
-                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserSemanticRoleLabelerAE</annotatorImplementationName>
+                <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPSemanticRoleLabelerAE</annotatorImplementationName>
                 <analysisEngineMetaData>
-                    <name>ClearParserSemanticRoleLabelerAE</name>
+                    <name>ClearNLPSemanticRoleLabelerAE</name>
                     <description>Descriptor automatically generated by uimaFIT</description>
                     <version>unknown</version>
                     <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
                     <configurationParameters>
                         <configurationParameter>
                             <name>ParserModelFileName</name>
-                            <description>This parameter provides the file name of the semantic role labeler model required by the factory method provided by ClearParserUtil.</description>
+                            <description>This parameter provides the file name of the semantic role labeler model required by the factory method provided by ClearNLPUtil.</description>
                             <type>String</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
                         </configurationParameter>
                         <configurationParameter>
                             <name>UseLemmatizer</name>
-                            <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                            <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                             <type>Boolean</type>
                             <multiValued>false</multiValued>
                             <mandatory>false</mandatory>
@@ -522,8 +522,8 @@
                 <node>SimpleSegmentAnnotator</node>
                 <node>LVG Annotator</node>
                 <node>POSTagger</node>
-                <node>ClearParserDependencyParserAE</node>
-                <node>ClearParserSemanticRoleLabelerAE</node>
+                <node>ClearNLPDependencyParserAE</node>
+                <node>ClearNLPSemanticRoleLabelerAE</node>
             </fixedFlow>
         </flowConstraints>
         <capabilities/>

Copied: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSemanticRoleLabelerAE.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSemanticRoleLabelerAE.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSemanticRoleLabelerAE.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSemanticRoleLabelerAE.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSemanticRoleLabelerAE.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearParserSemanticRoleLabelerAE.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/ClearNLPSemanticRoleLabelerAE.xml Fri Apr  5 16:53:06 2013
@@ -22,23 +22,23 @@
 <analysisEngineDescription xmlns="http://uima.apache.org/resourceSpecifier">
     <frameworkImplementation>org.apache.uima.java</frameworkImplementation>
     <primitive>true</primitive>
-    <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearParserSemanticRoleLabelerAE</annotatorImplementationName>
+    <annotatorImplementationName>org.apache.ctakes.dependency.parser.ae.ClearNLPSemanticRoleLabelerAE</annotatorImplementationName>
     <analysisEngineMetaData>
-        <name>ClearParserSemanticRoleLabelerAE</name>
+        <name>ClearNLPSemanticRoleLabelerAE</name>
         <description>Descriptor automatically generated by uimaFIT</description>
         <version>unknown</version>
         <vendor>org.apache.ctakes.dependency.parser.ae</vendor>
         <configurationParameters>
             <configurationParameter>
                 <name>ParserModelFileName</name>
-                <description>This parameter provides the file name of the semantic role labeler model required by the factory method provided by ClearParserUtil.</description>
+                <description>This parameter provides the file name of the semantic role labeler model required by the factory method provided by ClearNLPUtil.</description>
                 <type>String</type>
                 <multiValued>false</multiValued>
                 <mandatory>false</mandatory>
             </configurationParameter>
             <configurationParameter>
                 <name>UseLemmatizer</name>
-                <description>If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
+                <description>If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field</description>
                 <type>Boolean</type>
                 <multiValued>false</multiValued>
                 <mandatory>false</mandatory>

Modified: ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/README.txt
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/README.txt?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/README.txt (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/analysis_engine/README.txt Fri Apr  5 16:53:06 2013
@@ -1,2 +1,2 @@
-Do not manually edit the descriptors in this directory.  They are automatically generated using by WriteClearParserDescriptors, 
+Do not manually edit the descriptors in this directory.  They are automatically generated using by WriteClearNLPDescriptors, 
 because these analysis engines are written in UIMAFit and are not of the usual UIMA variety.
\ No newline at end of file

Copied: ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearNLPTestCPE.xml (from r1464565, ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearParserTestCPE.xml)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearNLPTestCPE.xml?p2=ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearNLPTestCPE.xml&p1=ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearParserTestCPE.xml&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearParserTestCPE.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/desc/collection_processing_engine/ClearNLPTestCPE.xml Fri Apr  5 16:53:06 2013
@@ -42,9 +42,9 @@
         </collectionIterator>
     </collectionReader>
     <casProcessors casPoolSize="3" processingUnitThreadCount="1">
-        <casProcessor deployment="integrated" name="ClearParserDependencyParserTokenizedInfPosAggregate">
+        <casProcessor deployment="integrated" name="ClearNLPDependencyParserTokenizedInfPosAggregate">
             <descriptor>
-                <import location="../analysis_engine/ClearParserDependencyParserTokenizedInfPosAggregate.xml"/>
+                <import location="../analysis_engine/ClearNLPDependencyParserTokenizedInfPosAggregate.xml"/>
             </descriptor>
             <deploymentParameters/>
             <errorHandling>

Modified: ctakes/trunk/ctakes-dependency-parser/pom.xml
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/pom.xml?rev=1465043&r1=1465042&r2=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/pom.xml (original)
+++ ctakes/trunk/ctakes-dependency-parser/pom.xml Fri Apr  5 16:53:06 2013
@@ -28,13 +28,6 @@
 		<artifactId>ctakes</artifactId>
 		<version>3.1.0-SNAPSHOT</version>
 	</parent>
-	<!-- necessary for ClearParser -->
-	<repositories>
-		<repository>
-			<id>cleartk-googlecode</id>
-			<url>http://cleartk.googlecode.com/svn/repo/</url>
-		</repository>
-	</repositories>
 	<dependencies>
 		<dependency>
 			<groupId>org.apache.ctakes</groupId>
@@ -60,34 +53,14 @@
 			<groupId>org.apache.ctakes</groupId>
 			<artifactId>ctakes-utils</artifactId>
 		</dependency>
-		
 		<dependency>
-			<groupId>com.googlecode.clearparser</groupId>
-			<artifactId>clearparser</artifactId>
-			<version>0.4.0-SNAPSHOT</version>
-			<exclusions>
-				<exclusion>
-					<groupId>org.apache.commons</groupId>
-					<artifactId>commons-compress</artifactId>
-				</exclusion>
-				<exclusion>
-					<groupId>org.apache.opennlp</groupId>
-					<artifactId>opennlp-maxent</artifactId>
-				</exclusion>
-				<exclusion>
-					<groupId>org.apache.opennlp</groupId>
-					<artifactId>opennlp-tools</artifactId>
-				</exclusion>
-			</exclusions>
+			<groupId>com.googlecode.clearnlp</groupId>
+			<artifactId>clearnlp</artifactId>
 		</dependency>
-		
 		<dependency>
 			<groupId>args4j</groupId>
 			<artifactId>args4j</artifactId>
 		</dependency>
-		
-       
-        
 		<dependency>
 			<groupId>junit</groupId>
 			<artifactId>junit</artifactId>
@@ -117,5 +90,5 @@
 			<artifactId>commons-logging</artifactId>
 		</dependency>
 	</dependencies>
-	<description>This wraps the ClearParser dependency parser and semantic role labeler into a UIMA friendly annotator</description>
+	<description>This wraps the ClearNLP dependency parser and semantic role labeler into a UIMA friendly annotator</description>
 </project>

Copied: ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearNLPDependencyParserAE.java (from r1464565, ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearParserDependencyParserAE.java)
URL: http://svn.apache.org/viewvc/ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearNLPDependencyParserAE.java?p2=ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearNLPDependencyParserAE.java&p1=ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearParserDependencyParserAE.java&r1=1464565&r2=1465043&rev=1465043&view=diff
==============================================================================
--- ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearParserDependencyParserAE.java (original)
+++ ctakes/trunk/ctakes-dependency-parser/src/main/java/org/apache/ctakes/dependency/parser/ae/ClearNLPDependencyParserAE.java Fri Apr  5 16:53:06 2013
@@ -19,11 +19,14 @@
 package org.apache.ctakes.dependency.parser.ae;
 
 import java.io.File;
+import java.io.FileInputStream;
 import java.io.IOException;
 import java.net.MalformedURLException;
+import java.net.URI;
 import java.net.URL;
 import java.util.ArrayList;
 import java.util.List;
+import java.util.zip.ZipInputStream;
 
 import org.apache.ctakes.dependency.parser.util.ClearDependencyUtility;
 import org.apache.ctakes.typesystem.type.syntax.BaseToken;
@@ -33,26 +36,30 @@ import org.apache.log4j.Logger;
 import org.apache.uima.UimaContext;
 import org.apache.uima.analysis_engine.AnalysisEngineProcessException;
 import org.apache.uima.jcas.JCas;
-import org.apache.uima.resource.ResourceAccessException;
 import org.apache.uima.resource.ResourceInitializationException;
 import org.uimafit.component.JCasAnnotator_ImplBase;
 import org.uimafit.descriptor.ConfigurationParameter;
 import org.uimafit.descriptor.TypeCapability;
 import org.uimafit.util.JCasUtil;
 
-import clear.dep.DepNode;
-import clear.dep.DepTree;
-import clear.morph.MorphEnAnalyzer;
-import clear.parse.AbstractDepParser;
+import com.googlecode.clearnlp.component.AbstractComponent;
+import com.googlecode.clearnlp.component.morph.CEnglishMPAnalyzer;
+import com.googlecode.clearnlp.dependency.DEPFeat;
+import com.googlecode.clearnlp.dependency.DEPNode;
+import com.googlecode.clearnlp.dependency.DEPTree;
+import com.googlecode.clearnlp.engine.EngineGetter;
+import com.googlecode.clearnlp.morphology.AbstractMPAnalyzer;
+import com.googlecode.clearnlp.nlp.NLPLib;
+import com.googlecode.clearnlp.reader.AbstractReader;
 
 /**
  * <br>
 * This class provides a UIMA wrapper for the CLEAR dependency parser. This parser is available here:
  * <p>
- * http://code.google.com/p/clearparser/
+ * http://code.google.com/p/clearnlp
  * <p>
  * Please see
- * /clearparser-wrapper/resources/dependency/README
+ * /ClearNLP-wrapper/resources/dependency/README
  * for important information pertaining to the models provided for this parser. In particular, note
  * that the output of the CLEAR parser is different than that of the Malt parser and so these two
  * parsers may not be interchangeable (without some effort) for most use cases.
@@ -68,17 +75,14 @@ import clear.parse.AbstractDepParser;
 				"org.apache.ctakes.typesystem.type.syntax.BaseToken:end",
 				"org.apache.ctakes.typesystem.type.syntax.BaseToken:begin"
 		})
-public class ClearParserDependencyParserAE extends JCasAnnotator_ImplBase {
-
+public class ClearNLPDependencyParserAE extends JCasAnnotator_ImplBase {
+
+	final String language = AbstractReader.LANG_EN;
 	public Logger logger = Logger.getLogger(getClass().getName());
 	
 	// Default model values
-	public static final String DEFAULT_MODEL_FILE_NAME = "org/apache/ctakes/dependency/parser/models/dependency/conll-2009-dev-shift-pop.jar";
-
-	public static final String ENG_LEMMATIZER_DATA_FILE = "org/apache/ctakes/dependency/parser/models/lemmatizer/wordnet-3.0-lemma-data.jar";
-
-	public static final String DEFAULT_PARSER_ALGORITHM_NAME = AbstractDepParser.ALG_SHIFT_POP;
-
+	public static final String DEFAULT_MODEL_FILE_NAME = "org/apache/ctakes/dependency/parser/models/dependency/mayo-en-dep-1.3.0.jar";
+	public static final String ENG_LEMMATIZER_DATA_FILE = "org/apache/ctakes/dependency/parser/models/lemmatizer/dictionary-1.3.1.jar";
 	
 	
 	// Configuration Parameters 
@@ -86,9 +90,9 @@ public class ClearParserDependencyParser
 	@ConfigurationParameter(
 			name = PARAM_PARSER_MODEL_FILE_NAME,
 			description = "This parameter provides the file name of the dependency parser model required " +
-					      "by the factory method provided by ClearParserUtil.  If not specified, this " +
+					      "by the factory method provided by ClearNLPUtil.  If not specified, this " +
 					      "analysis engine will use a default model from the resources directory")
-	protected String parserModelFileName;
+	protected URI parserModelUri;
 
   public static final String PARAM_LEMMATIZER_DATA_FILE = "LemmatizerDataFile";
 
@@ -96,28 +100,18 @@ public class ClearParserDependencyParser
       name = PARAM_LEMMATIZER_DATA_FILE,
       description = "This parameter provides the data file required for the MorphEnAnalyzer. If not "
           + "specified, this analysis engine will use a default model from the resources directory")
-  protected File lemmatizerDataFile;
+  protected URI lemmatizerDataFile;
 
-	public static final String PARAM_PARSER_ALGORITHM_NAME = "ParserAlgorithmName";
-	@ConfigurationParameter(
-			name = PARAM_PARSER_ALGORITHM_NAME,
-			defaultValue = DEFAULT_PARSER_ALGORITHM_NAME,
-			mandatory = true,
-			description = "This parameter provides the algorithm name used by the dependency parser that " +
-					      "is required by the factory method provided by ClearParserUtil.  " +
-					      "If in doubt, do not change from the default value.")
-	protected String parserAlgorithmName;
-
 	public static final String PARAM_USE_LEMMATIZER = "UseLemmatizer";
 	@ConfigurationParameter(
 			name = PARAM_USE_LEMMATIZER,
 			defaultValue = "true",
-			description = "If true, use the default ClearParser lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field")
+			description = "If true, use the default ClearNLP lemmatizer, otherwise use lemmas from the BaseToken normalizedToken field")
 	protected boolean useLemmatizer;
 
 
-	protected AbstractDepParser parser;
-	protected MorphEnAnalyzer lemmatizer;
+	protected AbstractComponent parser;
+	protected AbstractMPAnalyzer lemmatizer;
 	//protected boolean useLemmatizer = false;
 
 	@Override
@@ -130,44 +124,40 @@ public class ClearParserDependencyParser
 			if (useLemmatizer) {
 				// Note: If lemmatizer data file is not specified, then use lemmas from the BaseToken normalizedToken field.
 				// Initialize lemmatizer
-        URL lemmatizerDataFileURL = this.lemmatizerDataFile == null
-            ? this.getClass().getClassLoader().getResource(ENG_LEMMATIZER_DATA_FILE)
-            : this.lemmatizerDataFile.toURI().toURL();
-				lemmatizer = new MorphEnAnalyzer(lemmatizerDataFileURL);
-			}
-
-			// Initialize parser
-			URL parserModelURL = this.parserModelFileName == null
-					? this.getClass().getClassLoader().getResource(DEFAULT_MODEL_FILE_NAME)
-					: new File(this.parserModelFileName).toURI().toURL();
-			parser = ClearParserUtil.createParser(parserModelURL.openStream(), parserAlgorithmName);
-		} catch (MalformedURLException e) {
-			throw new ResourceInitializationException(e);
-		} catch (IOException e) {
-			throw new ResourceInitializationException(e);
-		}
-
+				
+                URL lemmatizerModelURL = (this.lemmatizerDataFile == null)
+                        ? this.getClass().getClassLoader().getResource(ENG_LEMMATIZER_DATA_FILE).toURI().toURL()
+                        : this.lemmatizerDataFile.toURL();
+                        
+                    this.lemmatizer = EngineGetter.getMPAnalyzer(language, lemmatizerModelURL.openStream());
+			}
+            	URL parserModelURL = (this.parserModelUri == null)
+                    ? this.getClass().getClassLoader().getResource(DEFAULT_MODEL_FILE_NAME).toURI().toURL()
+                    : this.parserModelUri.toURL();
+                 
+                    this.parser = EngineGetter.getComponent(parserModelURL.openStream(), this.language, NLPLib.MODE_DEP);
+
+        } catch (Exception e) {
+            throw new ResourceInitializationException(e);
+        }
 	}
 
 	@Override
 	public void process(JCas jCas) throws AnalysisEngineProcessException {
 		for (Sentence sentence : JCasUtil.select(jCas, Sentence.class)) {
 			List<BaseToken> tokens = JCasUtil.selectCovered(jCas, BaseToken.class, sentence);
-			DepTree tree = new DepTree();
+			DEPTree tree = new DEPTree();
 
-			// Convert CAS data into structures usable by ClearParser
+			// Convert CAS data into structures usable by ClearNLP
 			for (int i = 0; i < tokens.size(); i++) {
-				BaseToken token = tokens.get(i);
-				DepNode node = new DepNode();
-				node.id = i + 1;
-				node.form = token.getCoveredText();
-				node.pos = token.getPartOfSpeech();
-				node.lemma = useLemmatizer ? lemmatizer.getLemma(node.form, node.pos) : token.getNormalizedForm();
+				BaseToken token = tokens.get(i);
+				String lemma = useLemmatizer ? lemmatizer.getLemma(token.getCoveredText(), token.getPartOfSpeech()) : token.getNormalizedForm();
+				DEPNode node = new DEPNode(i+1, token.getCoveredText(), lemma, token.getPartOfSpeech(), new DEPFeat());
 				tree.add(node);
 			}
 
 			// Run parser and convert output back to CAS friendly data types
-			parser.parse(tree);
+			parser.process(tree);
 			ArrayList<ConllDependencyNode> nodes = ClearDependencyUtility.convert(jCas, tree, sentence, tokens);
 			ClearDependencyUtility.addToIndexes(jCas, nodes);
 		}



Mime
View raw message