Author: joern
Date: Wed Apr 15 15:19:32 2015
New Revision: 1673824
URL: http://svn.apache.org/r1673824
Log:
OPENNLP-766 Added automated name finder evaluation test using CONLL 2002 data
Added:
opennlp/trunk/opennlp-tools/src/test/java/opennlp/tools/eval/Conll02NameFinderEval.java
Added:
opennlp/trunk/opennlp-tools/src/test/java/opennlp/tools/eval/Conll02NameFinderEval.java
URL:
http://svn.apache.org/viewvc/opennlp/trunk/opennlp-tools/src/test/java/opennlp/tools/eval/Conll02NameFinderEval.java?rev=1673824&view=auto
==============================================================================
---
opennlp/trunk/opennlp-tools/src/test/java/opennlp/tools/eval/Conll02NameFinderEval.java
(added)
+++
opennlp/trunk/opennlp-tools/src/test/java/opennlp/tools/eval/Conll02NameFinderEval.java
Wed Apr 15 15:19:32 2015
@@ -0,0 +1,231 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+
+package opennlp.tools.eval;
+
+import java.io.File;
+import java.io.IOException;
+
+import opennlp.tools.formats.Conll02NameSampleStream;
+import opennlp.tools.formats.Conll02NameSampleStream.LANGUAGE;
+import opennlp.tools.namefind.NameFinderME;
+import opennlp.tools.namefind.NameSample;
+import opennlp.tools.namefind.TokenNameFinderEvaluator;
+import opennlp.tools.namefind.TokenNameFinderFactory;
+import opennlp.tools.namefind.TokenNameFinderModel;
+import opennlp.tools.util.MarkableFileInputStreamFactory;
+import opennlp.tools.util.ObjectStream;
+import opennlp.tools.util.TrainingParameters;
+import opennlp.tools.util.model.ModelUtil;
+
+import org.junit.Assert;
+import org.junit.Test;
+
+/**
+ * Evaluates the name finder against the Dutch and Spanish CONLL2002 corpus.
+ * <p>
+ * Download the tarball from the CONLL2002 shared task
+ * <a href="http://www.cnts.ua.ac.be/conll2002/ner/";> site </a>
+ * and decompress it into this directory: $OPENNLP_DATA_DIR/conll2002.
+ * Also decompress the training files.
+ *
+ * TODO:
+ * - Files are provided in gzipped. It would be better if they would not be
unpacked by the user.
+ * - Double check the encoding which is used to open the files. Currently that
is UTF-8.
+ * - Make the Conll02 reader compatible. Currently it doesn't work with
spanish data without pos tags.
+ */
+public class Conll02NameFinderEval {
+
+ private static TokenNameFinderModel train(File trainFile, LANGUAGE lang,
+ TrainingParameters params, int types) throws IOException {
+
+ ObjectStream<NameSample> samples = new Conll02NameSampleStream(
+ lang,new MarkableFileInputStreamFactory(trainFile), types);
+
+ return NameFinderME.train(lang.toString().toLowerCase(), null, samples,
+ params, new TokenNameFinderFactory());
+ }
+
+ private static void eval(TokenNameFinderModel model, File testData, LANGUAGE
lang,
+ int types, double expectedFMeasure) throws IOException {
+
+ ObjectStream<NameSample> samples = new Conll02NameSampleStream(
+ lang, new MarkableFileInputStreamFactory(testData), types);
+
+ TokenNameFinderEvaluator evaluator = new TokenNameFinderEvaluator(new
NameFinderME(model));
+ evaluator.evaluate(samples);
+
+ Assert.assertEquals(expectedFMeasure,
evaluator.getFMeasure().getFMeasure(), 0.0001);
+ }
+
+ @Test
+ public void evalDutchPerson() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.train"), LANGUAGE.NL, params,
+ Conll02NameSampleStream.GENERATE_PERSON_ENTITIES);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testa"), LANGUAGE.NL,
Conll02NameSampleStream.GENERATE_PERSON_ENTITIES, 0.5696539485359361d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testb"), LANGUAGE.NL,
Conll02NameSampleStream.GENERATE_PERSON_ENTITIES, 0.7127771911298839d);
+ }
+
+ @Test
+ public void evalDutchOrganization() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.train"), LANGUAGE.NL, params,
+ Conll02NameSampleStream.GENERATE_ORGANIZATION_ENTITIES);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testa"), LANGUAGE.NL,
Conll02NameSampleStream.GENERATE_ORGANIZATION_ENTITIES, 0.5197969543147207d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testb"), LANGUAGE.NL,
Conll02NameSampleStream.GENERATE_ORGANIZATION_ENTITIES, 0.5753228120516498d);
+ }
+
+ @Test
+ public void evalDutchLocation() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.train"), LANGUAGE.NL, params,
+ Conll02NameSampleStream.GENERATE_LOCATION_ENTITIES);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testa"), LANGUAGE.NL,
Conll02NameSampleStream.GENERATE_LOCATION_ENTITIES, 0.5451977401129944d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testb"), LANGUAGE.NL,
Conll02NameSampleStream.GENERATE_LOCATION_ENTITIES, 0.680952380952381d);
+ }
+
+ @Test
+ public void evalDutchMisc() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.train"), LANGUAGE.NL, params,
+ Conll02NameSampleStream.GENERATE_MISC_ENTITIES);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testa"), LANGUAGE.NL,
Conll02NameSampleStream.GENERATE_MISC_ENTITIES, 0.5831157528285466d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testb"), LANGUAGE.NL,
Conll02NameSampleStream.GENERATE_MISC_ENTITIES, 0.5762897914379803d);
+ }
+
+ @Test
+ public void evalDutchCombined() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ int combinedType = Conll02NameSampleStream.GENERATE_PERSON_ENTITIES |
Conll02NameSampleStream.GENERATE_ORGANIZATION_ENTITIES
+ | Conll02NameSampleStream.GENERATE_LOCATION_ENTITIES |
Conll02NameSampleStream.GENERATE_MISC_ENTITIES;
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.train"), LANGUAGE.NL, params,
+ combinedType);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testa"), LANGUAGE.NL, combinedType,
0.6728164867517175d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/ned.testb"), LANGUAGE.NL, combinedType,
0.6985893619774816d);
+ }
+
+ @Test
+ public void evalSpanishPerson() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.train"), LANGUAGE.ES, params,
+ Conll02NameSampleStream.GENERATE_PERSON_ENTITIES);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testa"), LANGUAGE.ES,
Conll02NameSampleStream.GENERATE_PERSON_ENTITIES, 0.686960933536276d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testb"), LANGUAGE.ES,
Conll02NameSampleStream.GENERATE_PERSON_ENTITIES, 0.8132033008252063d);
+ }
+
+ @Test
+ public void evalSpanishOrganization() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.train"), LANGUAGE.ES, params,
+ Conll02NameSampleStream.GENERATE_ORGANIZATION_ENTITIES);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testa"), LANGUAGE.ES,
Conll02NameSampleStream.GENERATE_ORGANIZATION_ENTITIES, 0.6982288828337874d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testb"), LANGUAGE.ES,
Conll02NameSampleStream.GENERATE_ORGANIZATION_ENTITIES, 0.7640449438202247d);
+ }
+
+ @Test
+ public void evalSpanishLocation() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.train"), LANGUAGE.ES, params,
+ Conll02NameSampleStream.GENERATE_LOCATION_ENTITIES);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testa"), LANGUAGE.ES,
Conll02NameSampleStream.GENERATE_LOCATION_ENTITIES, 0.7386907929749867d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testb"), LANGUAGE.ES,
Conll02NameSampleStream.GENERATE_LOCATION_ENTITIES, 0.6772777167947311d);
+ }
+
+ @Test
+ public void evalSpanishMisc() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.train"), LANGUAGE.ES, params,
+ Conll02NameSampleStream.GENERATE_MISC_ENTITIES);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testa"), LANGUAGE.ES,
Conll02NameSampleStream.GENERATE_MISC_ENTITIES, 0.40971168437025796d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testb"), LANGUAGE.ES,
Conll02NameSampleStream.GENERATE_MISC_ENTITIES, 0.45703124999999994d);
+ }
+
+ @Test
+ public void evalSpanishCombined() throws IOException {
+ TrainingParameters params = ModelUtil.createDefaultTrainingParameters();
+
+ int combinedType = Conll02NameSampleStream.GENERATE_PERSON_ENTITIES |
Conll02NameSampleStream.GENERATE_ORGANIZATION_ENTITIES
+ | Conll02NameSampleStream.GENERATE_LOCATION_ENTITIES |
Conll02NameSampleStream.GENERATE_MISC_ENTITIES;
+
+ TokenNameFinderModel maxentModel = train(new
File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.train"), LANGUAGE.ES, params,
+ combinedType);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testa"), LANGUAGE.ES, combinedType,
0.706765154179857d);
+
+ eval(maxentModel, new File(EvalUtil.getOpennlpDataDir(),
+ "conll02/ner/data/esp.testb"), LANGUAGE.ES, combinedType,
0.7583580194667795d);
+ }
+}
