From 385cf9ed1d2462d4ced2b7cdffccc5df2dc7ca3d Mon Sep 17 00:00:00 2001
From: dlutz2 <dlutz@mitre.org>
Date: Mon, 26 Sep 2016 09:43:03 -0400
Subject: [PATCH] Switching to Emory lemmatizer

---
 pom.xml                                        |  8 ++++----
 src/org/opensextant/howler/utils/OWLUtils.java | 15 ++++++++++++---
 2 files changed, 16 insertions(+), 7 deletions(-)

diff --git a/pom.xml b/pom.xml
index c9aa1c0..f02f976 100644
--- a/pom.xml
+++ b/pom.xml
@@ -109,11 +109,11 @@
 			<version>1.7.14</version>
 		</dependency>
 
-		<!-- Stanford NLP core, only word normalizer used -->
+		<!-- Emory Lemmatizer -->
 		<dependency>
-			<groupId>edu.stanford.nlp</groupId>
-			<artifactId>stanford-corenlp</artifactId>
-			<version>3.6.0</version>
+			<groupId>edu.emory.mathcs.nlp</groupId>
+			<artifactId>nlp4j-morphology</artifactId>
+			<version>1.1.2</version>
 		</dependency>
 
 		<!-- Part of speech tagger, used in Lexer -->
diff --git a/src/org/opensextant/howler/utils/OWLUtils.java b/src/org/opensextant/howler/utils/OWLUtils.java
index fb03991..306d6ec 100644
--- a/src/org/opensextant/howler/utils/OWLUtils.java
+++ b/src/org/opensextant/howler/utils/OWLUtils.java
@@ -45,12 +45,16 @@
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
-import edu.stanford.nlp.process.Morphology;
+import edu.emory.mathcs.nlp.common.util.StringUtils;
+import edu.emory.mathcs.nlp.component.morph.MorphAnalyzer;
+import edu.emory.mathcs.nlp.component.morph.english.EnglishMorphAnalyzer;
 
 public class OWLUtils {
 
   static Map<String, Integer> numbers = new HashMap<String, Integer>();
-
+  
+  static MorphAnalyzer lemmatizer = new EnglishMorphAnalyzer();
+  
   static {
     numbers.put("one", 1);
     numbers.put("two", 2);
@@ -199,12 +203,17 @@ public static String normalize(String word, String pos, boolean lower) {
       return word;
     }
 
+    // don't change numbers or Fixed vocab
+    if (pos.equals("CD") || pos.equals("FIXED")) {
+      return word;
+    }
+
     // don't normalize verbs yet
     if (pos.startsWith("V")) {
       return word;
     }
 
-    return (Morphology.lemmaStatic(word, pos, lower));
+    return lemmatizer.lemmatize(StringUtils.toSimplifiedForm(word, lower), pos);
   }
 
   public static SubjectPredicateObject rewriteSPO(