You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@lucene.apache.org by to...@apache.org on 2017/04/07 08:59:41 UTC
[1/3] lucene-solr:master: LUCENE-5548 - improved testing for SNBC
Repository: lucene-solr
Updated Branches:
refs/heads/master c05ab96dc -> f37fad206
LUCENE-5548 - improved testing for SNBC
Project: http://git-wip-us.apache.org/repos/asf/lucene-solr/repo
Commit: http://git-wip-us.apache.org/repos/asf/lucene-solr/commit/f37fad20
Tree: http://git-wip-us.apache.org/repos/asf/lucene-solr/tree/f37fad20
Diff: http://git-wip-us.apache.org/repos/asf/lucene-solr/diff/f37fad20
Branch: refs/heads/master
Commit: f37fad206b0dbf32209bff3761179458a1ddf7bf
Parents: 276ccff
Author: Tommaso Teofili <to...@apache.org>
Authored: Fri Apr 7 10:58:49 2017 +0200
Committer: Tommaso Teofili <to...@apache.org>
Committed: Fri Apr 7 10:59:20 2017 +0200
----------------------------------------------------------------------
.../classification/SimpleNaiveBayesClassifierTest.java | 11 +++++++++--
1 file changed, 9 insertions(+), 2 deletions(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/lucene-solr/blob/f37fad20/lucene/classification/src/test/org/apache/lucene/classification/SimpleNaiveBayesClassifierTest.java
----------------------------------------------------------------------
diff --git a/lucene/classification/src/test/org/apache/lucene/classification/SimpleNaiveBayesClassifierTest.java b/lucene/classification/src/test/org/apache/lucene/classification/SimpleNaiveBayesClassifierTest.java
index 2b4873d..0e05d4f 100644
--- a/lucene/classification/src/test/org/apache/lucene/classification/SimpleNaiveBayesClassifierTest.java
+++ b/lucene/classification/src/test/org/apache/lucene/classification/SimpleNaiveBayesClassifierTest.java
@@ -59,8 +59,10 @@ public class SimpleNaiveBayesClassifierTest extends ClassificationTestBase<Bytes
try {
MockAnalyzer analyzer = new MockAnalyzer(random());
leafReader = getSampleIndex(analyzer);
- TermQuery query = new TermQuery(new Term(textFieldName, "it"));
- checkCorrectClassification(new SimpleNaiveBayesClassifier(leafReader, analyzer, query, categoryFieldName, textFieldName), TECHNOLOGY_INPUT, TECHNOLOGY_RESULT);
+ TermQuery query = new TermQuery(new Term(textFieldName, "a"));
+ SimpleNaiveBayesClassifier classifier = new SimpleNaiveBayesClassifier(leafReader, analyzer, query, categoryFieldName, textFieldName);
+ checkCorrectClassification(classifier, TECHNOLOGY_INPUT, TECHNOLOGY_RESULT);
+ checkCorrectClassification(classifier, POLITICS_INPUT, POLITICS_RESULT);
} finally {
if (leafReader != null) {
leafReader.close();
@@ -112,6 +114,11 @@ public class SimpleNaiveBayesClassifierTest extends ClassificationTestBase<Bytes
assertTrue("evaluation took more than 2m: " + evaluationTime / 1000 + "s", evaluationTime < 120000);
double avgClassificationTime = confusionMatrix.getAvgClassificationTime();
assertTrue("avg classification time: " + avgClassificationTime, 5000 > avgClassificationTime);
+
+ double f1 = confusionMatrix.getF1Measure();
+ assertTrue(f1 >= 0d);
+ assertTrue(f1 <= 1d);
+
double accuracy = confusionMatrix.getAccuracy();
assertTrue(accuracy >= 0d);
assertTrue(accuracy <= 1d);
[3/3] lucene-solr:master: LUCENE-6853 - renamed threshold to bias,
initialize to avg tf
Posted by to...@apache.org.
LUCENE-6853 - renamed threshold to bias, initialize to avg tf
Project: http://git-wip-us.apache.org/repos/asf/lucene-solr/repo
Commit: http://git-wip-us.apache.org/repos/asf/lucene-solr/commit/cbad533d
Tree: http://git-wip-us.apache.org/repos/asf/lucene-solr/tree/cbad533d
Diff: http://git-wip-us.apache.org/repos/asf/lucene-solr/diff/cbad533d
Branch: refs/heads/master
Commit: cbad533d7a44a5fd41f85756c791f3d7439861a2
Parents: c05ab96
Author: Tommaso Teofili <to...@apache.org>
Authored: Thu Apr 6 19:05:52 2017 +0200
Committer: Tommaso Teofili <to...@apache.org>
Committed: Fri Apr 7 10:59:20 2017 +0200
----------------------------------------------------------------------
.../BooleanPerceptronClassifier.java | 26 ++++++++++----------
.../BooleanPerceptronClassifierTest.java | 14 +++++++----
2 files changed, 22 insertions(+), 18 deletions(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/lucene-solr/blob/cbad533d/lucene/classification/src/java/org/apache/lucene/classification/BooleanPerceptronClassifier.java
----------------------------------------------------------------------
diff --git a/lucene/classification/src/java/org/apache/lucene/classification/BooleanPerceptronClassifier.java b/lucene/classification/src/java/org/apache/lucene/classification/BooleanPerceptronClassifier.java
index 781a14f..928c036 100644
--- a/lucene/classification/src/java/org/apache/lucene/classification/BooleanPerceptronClassifier.java
+++ b/lucene/classification/src/java/org/apache/lucene/classification/BooleanPerceptronClassifier.java
@@ -58,7 +58,7 @@ import org.apache.lucene.util.fst.Util;
*/
public class BooleanPerceptronClassifier implements Classifier<Boolean> {
- private final Double threshold;
+ private final Double bias;
private final Terms textTerms;
private final Analyzer analyzer;
private final String textFieldName;
@@ -72,14 +72,14 @@ public class BooleanPerceptronClassifier implements Classifier<Boolean> {
* @param query a {@link Query} to eventually filter the docs used for training the classifier, or {@code null}
* if all the indexed docs should be used
* @param batchSize the size of the batch of docs to use for updating the perceptron weights
- * @param threshold the threshold used for class separation
+ * @param bias the bias used for class separation
* @param classFieldName the name of the field used as the output for the classifier
* @param textFieldName the name of the field used as input for the classifier
* @throws IOException if the building of the underlying {@link FST} fails and / or {@link TermsEnum} for the text field
* cannot be found
*/
public BooleanPerceptronClassifier(IndexReader indexReader, Analyzer analyzer, Query query, Integer batchSize,
- Double threshold, String classFieldName, String textFieldName) throws IOException {
+ Double bias, String classFieldName, String textFieldName) throws IOException {
this.textTerms = MultiFields.getTerms(indexReader, textFieldName);
if (textTerms == null) {
@@ -89,18 +89,18 @@ public class BooleanPerceptronClassifier implements Classifier<Boolean> {
this.analyzer = analyzer;
this.textFieldName = textFieldName;
- if (threshold == null || threshold == 0d) {
- // automatic assign a threshold
- long sumDocFreq = indexReader.getSumDocFreq(textFieldName);
- if (sumDocFreq != -1) {
- this.threshold = (double) sumDocFreq / 2d;
+ if (bias == null || bias == 0d) {
+ // automatic assign the bias to be the average total term freq
+ double t = (double) indexReader.getSumTotalTermFreq(textFieldName) / (double) indexReader.getDocCount(textFieldName);
+ if (t != -1) {
+ this.bias = t;
} else {
throw new IOException(
- "threshold cannot be assigned since term vectors for field "
+ "bias cannot be assigned since term vectors for field "
+ textFieldName + " do not exist");
}
} else {
- this.threshold = threshold;
+ this.bias = bias;
}
// TODO : remove this map as soon as we have a writable FST
@@ -173,7 +173,7 @@ public class BooleanPerceptronClassifier implements Classifier<Boolean> {
// update weights
Long previousValue = Util.get(fst, term);
String termString = term.utf8ToString();
- weights.put(termString, previousValue == null ? 0 : previousValue + modifier * termFreqLocal);
+ weights.put(termString, previousValue == null ? 0 : Math.max(0, previousValue + modifier * termFreqLocal));
}
}
if (updateFST) {
@@ -216,8 +216,8 @@ public class BooleanPerceptronClassifier implements Classifier<Boolean> {
tokenStream.end();
}
- double score = 1 - Math.exp(-1 * Math.abs(threshold - output.doubleValue()) / threshold);
- return new ClassificationResult<>(output >= threshold, score);
+ double score = 1 - Math.exp(-1 * Math.abs(bias - output.doubleValue()) / bias);
+ return new ClassificationResult<>(output >= bias, score);
}
/**
http://git-wip-us.apache.org/repos/asf/lucene-solr/blob/cbad533d/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java
----------------------------------------------------------------------
diff --git a/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java b/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java
index 6ea92c0..ec059f7 100644
--- a/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java
+++ b/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java
@@ -34,7 +34,9 @@ public class BooleanPerceptronClassifierTest extends ClassificationTestBase<Bool
try {
MockAnalyzer analyzer = new MockAnalyzer(random());
leafReader = getSampleIndex(analyzer);
- checkCorrectClassification(new BooleanPerceptronClassifier(leafReader, analyzer, null, 1, null, booleanFieldName, textFieldName), TECHNOLOGY_INPUT, false);
+ BooleanPerceptronClassifier classifier = new BooleanPerceptronClassifier(leafReader, analyzer, null, 1, null, booleanFieldName, textFieldName);
+ checkCorrectClassification(classifier, TECHNOLOGY_INPUT, false);
+ checkCorrectClassification(classifier, POLITICS_INPUT, true);
} finally {
if (leafReader != null) {
leafReader.close();
@@ -60,12 +62,14 @@ public class BooleanPerceptronClassifierTest extends ClassificationTestBase<Bool
@Test
public void testBasicUsageWithQuery() throws Exception {
- TermQuery query = new TermQuery(new Term(textFieldName, "it"));
+ TermQuery query = new TermQuery(new Term(textFieldName, "of"));
LeafReader leafReader = null;
try {
MockAnalyzer analyzer = new MockAnalyzer(random());
leafReader = getSampleIndex(analyzer);
- checkCorrectClassification(new BooleanPerceptronClassifier(leafReader, analyzer, query, 1, null, booleanFieldName, textFieldName), TECHNOLOGY_INPUT, false);
+ BooleanPerceptronClassifier classifier = new BooleanPerceptronClassifier(leafReader, analyzer, query, 1, null, booleanFieldName, textFieldName);
+ checkCorrectClassification(classifier, TECHNOLOGY_INPUT, false);
+ checkCorrectClassification(classifier, POLITICS_INPUT, true);
} finally {
if (leafReader != null) {
leafReader.close();
@@ -94,8 +98,8 @@ public class BooleanPerceptronClassifierTest extends ClassificationTestBase<Bool
double avgClassificationTime = confusionMatrix.getAvgClassificationTime();
assertTrue(5000 > avgClassificationTime);
// accuracy check disabled until LUCENE-6853 is fixed
-// double accuracy = confusionMatrix.getAccuracy();
-// assertTrue(accuracy > 0d);
+ double accuracy = confusionMatrix.getAccuracy();
+ assertTrue(accuracy > 0d);
} finally {
leafReader.close();
}
[2/3] lucene-solr:master: LUCENE-6853 - re-enabled test
classification measures for bpc
Posted by to...@apache.org.
LUCENE-6853 - re-enabled test classification measures for bpc
Project: http://git-wip-us.apache.org/repos/asf/lucene-solr/repo
Commit: http://git-wip-us.apache.org/repos/asf/lucene-solr/commit/276ccff7
Tree: http://git-wip-us.apache.org/repos/asf/lucene-solr/tree/276ccff7
Diff: http://git-wip-us.apache.org/repos/asf/lucene-solr/diff/276ccff7
Branch: refs/heads/master
Commit: 276ccff751dc0f5cd0599dd5a6682553a0b58d7b
Parents: cbad533
Author: Tommaso Teofili <to...@apache.org>
Authored: Thu Apr 6 19:13:50 2017 +0200
Committer: Tommaso Teofili <to...@apache.org>
Committed: Fri Apr 7 10:59:20 2017 +0200
----------------------------------------------------------------------
.../BooleanPerceptronClassifierTest.java | 39 ++++++++++++++++++--
1 file changed, 36 insertions(+), 3 deletions(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/lucene-solr/blob/276ccff7/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java
----------------------------------------------------------------------
diff --git a/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java b/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java
index ec059f7..5ecf9c6 100644
--- a/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java
+++ b/lucene/classification/src/test/org/apache/lucene/classification/BooleanPerceptronClassifierTest.java
@@ -19,8 +19,12 @@ package org.apache.lucene.classification;
import org.apache.lucene.analysis.MockAnalyzer;
import org.apache.lucene.classification.utils.ConfusionMatrixGenerator;
import org.apache.lucene.index.LeafReader;
+import org.apache.lucene.index.MultiFields;
import org.apache.lucene.index.Term;
+import org.apache.lucene.index.Terms;
+import org.apache.lucene.index.TermsEnum;
import org.apache.lucene.search.TermQuery;
+import org.apache.lucene.util.BytesRef;
import org.junit.Test;
/**
@@ -90,16 +94,45 @@ public class BooleanPerceptronClassifierTest extends ClassificationTestBase<Bool
long evaluationStart = System.currentTimeMillis();
ConfusionMatrixGenerator.ConfusionMatrix confusionMatrix = ConfusionMatrixGenerator.getConfusionMatrix(leafReader,
- classifier, categoryFieldName, textFieldName, -1);
+ classifier, booleanFieldName, textFieldName, -1);
assertNotNull(confusionMatrix);
long evaluationEnd = System.currentTimeMillis();
long evaluationTime = evaluationEnd - evaluationStart;
assertTrue("evaluation took more than 1m: " + evaluationTime / 1000 + "s", evaluationTime < 60000);
double avgClassificationTime = confusionMatrix.getAvgClassificationTime();
assertTrue(5000 > avgClassificationTime);
- // accuracy check disabled until LUCENE-6853 is fixed
+
+ double f1 = confusionMatrix.getF1Measure();
+ assertTrue(f1 >= 0d);
+ assertTrue(f1 <= 1d);
+
double accuracy = confusionMatrix.getAccuracy();
- assertTrue(accuracy > 0d);
+ assertTrue(accuracy >= 0d);
+ assertTrue(accuracy <= 1d);
+
+ double recall = confusionMatrix.getRecall();
+ assertTrue(recall >= 0d);
+ assertTrue(recall <= 1d);
+
+ double precision = confusionMatrix.getPrecision();
+ assertTrue(precision >= 0d);
+ assertTrue(precision <= 1d);
+
+ Terms terms = MultiFields.getTerms(leafReader, booleanFieldName);
+ TermsEnum iterator = terms.iterator();
+ BytesRef term;
+ while ((term = iterator.next()) != null) {
+ String s = term.utf8ToString();
+ recall = confusionMatrix.getRecall(s);
+ assertTrue(recall >= 0d);
+ assertTrue(recall <= 1d);
+ precision = confusionMatrix.getPrecision(s);
+ assertTrue(precision >= 0d);
+ assertTrue(precision <= 1d);
+ double f1Measure = confusionMatrix.getF1Measure(s);
+ assertTrue(f1Measure >= 0d);
+ assertTrue(f1Measure <= 1d);
+ }
} finally {
leafReader.close();
}