基于用户词典的合并提前到紧接NGram切分后

hankcs · Jul 12, 2015 · b219c3a · b219c3a
1 parent 8d388ba
commit b219c3a
Show file tree

Hide file tree

Showing 3 changed files with 13 additions and 11 deletions.
diff --git a/src/main/java/com/hankcs/hanlp/seg/Dijkstra/DijkstraSegment.java b/src/main/java/com/hankcs/hanlp/seg/Dijkstra/DijkstraSegment.java
@@ -44,6 +44,12 @@ public List<Term> segSentence(char[] sentence)
         }
         List<Vertex> vertexList = dijkstra(graph);
 //        fixResultByRule(vertexList);
+
+        if (config.useCustomDictionary)
+        {
+            combineByCustomDictionary(vertexList);
+        }
+
         if (HanLP.Config.DEBUG)
         {
             System.out.println("粗分结果" + convert(vertexList, false));
@@ -110,11 +116,6 @@ public List<Term> segSentence(char[] sentence)
             speechTagging(vertexList);
         }
 
-        if (config.useCustomDictionary)
-        {
-            combineByCustomDictionary(vertexList);
-        }
-
         return convert(vertexList, config.offset);
     }
 

diff --git a/src/main/java/com/hankcs/hanlp/seg/Viterbi/ViterbiSegment.java b/src/main/java/com/hankcs/hanlp/seg/Viterbi/ViterbiSegment.java
@@ -49,6 +49,12 @@ protected List<Term> segSentence(char[] sentence)
 //        start = System.currentTimeMillis();
         List<Vertex> vertexList = viterbi(wordNetAll);
 //        System.out.println("最短路：" + (System.currentTimeMillis() - start));
+
+        if (config.useCustomDictionary)
+        {
+            combineByCustomDictionary(vertexList);
+        }
+
         if (HanLP.Config.DEBUG)
         {
             System.out.println("粗分结果" + convert(vertexList, false));
@@ -112,11 +118,6 @@ protected List<Term> segSentence(char[] sentence)
             speechTagging(vertexList);
         }
 
-        if (config.useCustomDictionary)
-        {
-            combineByCustomDictionary(vertexList);
-        }
-
         return convert(vertexList, config.offset);
     }
 

diff --git a/src/test/java/com/hankcs/test/seg/TestSegment.java b/src/test/java/com/hankcs/test/seg/TestSegment.java
@@ -55,7 +55,7 @@ public void testViterbi() throws Exception
         HanLP.Config.ShowTermNature = false;
         Segment segment = new DijkstraSegment();
         System.out.println(segment.seg(
-                "巴曙松：房地产如何估价才合理"
+                "上外日本文化经济学院的陆晚霞教授正在教授泛读课程"
         ));
     }