/*
* 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 org.apache.solr.update.processor;
import java.io.IOException;
import java.util.ArrayList;
import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.MockAnalyzer;
import org.apache.lucene.analysis.MockTokenizer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.RandomIndexWriter;
import org.apache.lucene.index.Term;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.TermQuery;
import org.apache.lucene.store.Directory;
import org.apache.solr.SolrTestCaseJ4;
import org.apache.solr.common.SolrInputDocument;
import org.apache.solr.update.AddUpdateCommand;
import org.junit.BeforeClass;
import org.junit.Test;
import static org.hamcrest.core.Is.is;
import static org.mockito.Mockito.mock;
/**
* Tests for {@link ClassificationUpdateProcessor}
*/
public class ClassificationUpdateProcessorTest extends SolrTestCaseJ4 {
/* field names are used in accordance with the solrconfig and schema supplied */
private static final String ID = "id";
private static final String TITLE = "title";
private static final String CONTENT = "content";
private static final String AUTHOR = "author";
private static final String TRAINING_CLASS = "cat";
private static final String PREDICTED_CLASS = "predicted";
public static final String KNN = "knn";
protected Directory directory;
protected IndexReader reader;
protected IndexSearcher searcher;
protected Analyzer analyzer = new MockAnalyzer(random(), MockTokenizer.WHITESPACE, false);
private ClassificationUpdateProcessor updateProcessorToTest;
@BeforeClass
public static void beforeClass() throws Exception {
System.setProperty("enable.update.log", "false");
initCore("solrconfig-classification.xml", "schema-classification.xml");
}
@Override
public void setUp() throws Exception {
super.setUp();
}
@Override
public void tearDown() throws Exception {
reader.close();
directory.close();
analyzer.close();
super.tearDown();
}
@Test
public void classificationMonoClass_predictedClassFieldSet_shouldAssignClassInPredictedClassField() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMonoClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word4 word4 word4",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params = initParams(ClassificationUpdateProcessorFactory.Algorithm.KNN);
params.setPredictedClassField(PREDICTED_CLASS);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
assertThat(unseenDocument1.getFieldValue(PREDICTED_CLASS),is("class2"));
}
@Test
public void knnMonoClass_sampleParams_shouldAssignCorrectClass() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMonoClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word4 word4 word4",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params = initParams(ClassificationUpdateProcessorFactory.Algorithm.KNN);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
assertThat(unseenDocument1.getFieldValue(TRAINING_CLASS),is("class2"));
}
@Test
public void knnMonoClass_boostFields_shouldAssignCorrectClass() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMonoClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word4 word4 word4",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params = initParams(ClassificationUpdateProcessorFactory.Algorithm.KNN);
params.setInputFieldNames(new String[]{TITLE + "^1.5", CONTENT + "^0.5", AUTHOR + "^2.5"});
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
assertThat(unseenDocument1.getFieldValue(TRAINING_CLASS),is("class2"));
}
@Test
public void bayesMonoClass_sampleParams_shouldAssignCorrectClass() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMonoClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word4 word4 word4",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params= initParams(ClassificationUpdateProcessorFactory.Algorithm.BAYES);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
assertThat(unseenDocument1.getFieldValue(TRAINING_CLASS),is("class1"));
}
@Test
public void knnMonoClass_contextQueryFiltered_shouldAssignCorrectClass() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMonoClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word4 word4 word4",
CONTENT, "word2 word2 ",
AUTHOR, "a");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params= initParams(ClassificationUpdateProcessorFactory.Algorithm.KNN);
Query class3DocsChunk=new TermQuery(new Term(TITLE,"word6"));
params.setTrainingFilterQuery(class3DocsChunk);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
assertThat(unseenDocument1.getFieldValue(TRAINING_CLASS),is("class3"));
}
@Test
public void bayesMonoClass_boostFields_shouldAssignCorrectClass() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMonoClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word4 word4 word4",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params= initParams(ClassificationUpdateProcessorFactory.Algorithm.BAYES);
params.setInputFieldNames(new String[]{TITLE+"^1.5",CONTENT+"^0.5",AUTHOR+"^2.5"});
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
assertThat(unseenDocument1.getFieldValue(TRAINING_CLASS),is("class2"));
}
@Test
public void knnClassification_maxOutputClassesGreaterThanAvailable_shouldAssignCorrectClass() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMultiClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word1 word1 word1",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params= initParams(ClassificationUpdateProcessorFactory.Algorithm.KNN);
params.setMaxPredictedClasses(100);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
ArrayList<Object> assignedClasses = (ArrayList)unseenDocument1.getFieldValues(TRAINING_CLASS);
assertThat(assignedClasses.get(0),is("class2"));
assertThat(assignedClasses.get(1),is("class1"));
}
@Test
public void knnMultiClass_maxOutputClasses2_shouldAssignMax2Classes() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMultiClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word1 word1 word1",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params= initParams(ClassificationUpdateProcessorFactory.Algorithm.KNN);
params.setMaxPredictedClasses(2);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
ArrayList<Object> assignedClasses = (ArrayList)unseenDocument1.getFieldValues(TRAINING_CLASS);
assertThat(assignedClasses.size(),is(2));
assertThat(assignedClasses.get(0),is("class2"));
assertThat(assignedClasses.get(1),is("class1"));
}
@Test
public void bayesMultiClass_maxOutputClasses2_shouldAssignMax2Classes() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMultiClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word1 word1 word1",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params= initParams(ClassificationUpdateProcessorFactory.Algorithm.BAYES);
params.setMaxPredictedClasses(2);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
ArrayList<Object> assignedClasses = (ArrayList)unseenDocument1.getFieldValues(TRAINING_CLASS);
assertThat(assignedClasses.size(),is(2));
assertThat(assignedClasses.get(0),is("class2"));
assertThat(assignedClasses.get(1),is("class1"));
}
@Test
public void knnMultiClass_boostFieldsMaxOutputClasses2_shouldAssignMax2Classes() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMultiClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word4 word4 word4",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params= initParams(ClassificationUpdateProcessorFactory.Algorithm.KNN);
params.setInputFieldNames(new String[]{TITLE+"^1.5",CONTENT+"^0.5",AUTHOR+"^2.5"});
params.setMaxPredictedClasses(2);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
ArrayList<Object> assignedClasses = (ArrayList)unseenDocument1.getFieldValues(TRAINING_CLASS);
assertThat(assignedClasses.size(),is(2));
assertThat(assignedClasses.get(0),is("class4"));
assertThat(assignedClasses.get(1),is("class6"));
}
@Test
public void bayesMultiClass_boostFieldsMaxOutputClasses2_shouldAssignMax2Classes() throws Exception {
UpdateRequestProcessor mockProcessor=mock(UpdateRequestProcessor.class);
prepareTrainedIndexMultiClass();
AddUpdateCommand update=new AddUpdateCommand(req());
SolrInputDocument unseenDocument1 = sdoc(ID, "10",
TITLE, "word4 word4 word4",
CONTENT, "word2 word2 ",
AUTHOR, "unseenAuthor");
update.solrDoc=unseenDocument1;
ClassificationUpdateProcessorParams params= initParams(ClassificationUpdateProcessorFactory.Algorithm.BAYES);
params.setInputFieldNames(new String[]{TITLE+"^1.5",CONTENT+"^0.5",AUTHOR+"^2.5"});
params.setMaxPredictedClasses(2);
updateProcessorToTest=new ClassificationUpdateProcessor(params,mockProcessor,reader,req().getSchema());
updateProcessorToTest.processAdd(update);
ArrayList<Object> assignedClasses = (ArrayList)unseenDocument1.getFieldValues(TRAINING_CLASS);
assertThat(assignedClasses.size(),is(2));
assertThat(assignedClasses.get(0),is("class4"));
assertThat(assignedClasses.get(1),is("class6"));
}
private ClassificationUpdateProcessorParams initParams(ClassificationUpdateProcessorFactory.Algorithm classificationAlgorithm) {
ClassificationUpdateProcessorParams params= new ClassificationUpdateProcessorParams();
params.setInputFieldNames(new String[]{TITLE,CONTENT,AUTHOR});
params.setTrainingClassField(TRAINING_CLASS);
params.setPredictedClassField(TRAINING_CLASS);
params.setMinTf(1);
params.setMinDf(1);
params.setK(5);
params.setAlgorithm(classificationAlgorithm);
params.setMaxPredictedClasses(1);
return params;
}
/**
* Index some example documents with a class manually assigned.
* This will be our trained model.
*
* @throws Exception If there is a low-level I/O error
*/
private void prepareTrainedIndexMonoClass() throws Exception {
directory = newDirectory();
RandomIndexWriter writer = new RandomIndexWriter(random(), directory);
//class1
addDoc(writer, buildLuceneDocument(ID, "1",
TITLE, "word1 word1 word1",
CONTENT, "word2 word2 word2",
AUTHOR, "a",
TRAINING_CLASS, "class1"));
addDoc(writer, buildLuceneDocument(ID, "2",
TITLE, "word1 word1",
CONTENT, "word2 word2",
AUTHOR, "a",
TRAINING_CLASS, "class1"));
addDoc(writer, buildLuceneDocument(ID, "3",
TITLE, "word1 word1 word1",
CONTENT, "word2",
AUTHOR, "a",
TRAINING_CLASS, "class1"));
addDoc(writer, buildLuceneDocument(ID, "4",
TITLE, "word1 word1 word1",
CONTENT, "word2 word2 word2",
AUTHOR, "a",
TRAINING_CLASS, "class1"));
//class2
addDoc(writer, buildLuceneDocument(ID, "5",
TITLE, "word4 word4 word4",
CONTENT, "word5 word5",
AUTHOR, "c",
TRAINING_CLASS, "class2"));
addDoc(writer, buildLuceneDocument(ID, "6",
TITLE, "word4 word4",
CONTENT, "word5",
AUTHOR, "c",
TRAINING_CLASS, "class2"));
addDoc(writer, buildLuceneDocument(ID, "7",
TITLE, "word4 word4 word4",
CONTENT, "word5 word5 word5",
AUTHOR, "c",
TRAINING_CLASS, "class2"));
addDoc(writer, buildLuceneDocument(ID, "8",
TITLE, "word4",
CONTENT, "word5 word5 word5 word5",
AUTHOR, "c",
TRAINING_CLASS, "class2"));
//class3
addDoc(writer, buildLuceneDocument(ID, "9",
TITLE, "word6",
CONTENT, "word7",
AUTHOR, "a",
TRAINING_CLASS, "class3"));
addDoc(writer, buildLuceneDocument(ID, "10",
TITLE, "word6",
CONTENT, "word7",
AUTHOR, "a",
TRAINING_CLASS, "class3"));
addDoc(writer, buildLuceneDocument(ID, "11",
TITLE, "word6",
CONTENT, "word7",
AUTHOR, "a",
TRAINING_CLASS, "class3"));
addDoc(writer, buildLuceneDocument(ID, "12",
TITLE, "word6",
CONTENT, "word7",
AUTHOR, "a",
TRAINING_CLASS, "class3"));
reader = writer.getReader();
writer.close();
searcher = newSearcher(reader);
}
private void prepareTrainedIndexMultiClass() throws Exception {
directory = newDirectory();
RandomIndexWriter writer = new RandomIndexWriter(random(), directory);
//class1
addDoc(writer, buildLuceneDocument(ID, "1",
TITLE, "word1 word1 word1",
CONTENT, "word2 word2 word2",
AUTHOR, "Name Surname",
TRAINING_CLASS, "class1",
TRAINING_CLASS, "class2"
));
addDoc(writer, buildLuceneDocument(ID, "2",
TITLE, "word1 word1",
CONTENT, "word2 word2",
AUTHOR, "Name Surname",
TRAINING_CLASS, "class3",
TRAINING_CLASS, "class2"
));
addDoc(writer, buildLuceneDocument(ID, "3",
TITLE, "word1 word1 word1",
CONTENT, "word2",
AUTHOR, "Name Surname",
TRAINING_CLASS, "class1",
TRAINING_CLASS, "class2"
));
addDoc(writer, buildLuceneDocument(ID, "4",
TITLE, "word1 word1 word1",
CONTENT, "word2 word2 word2",
AUTHOR, "Name Surname",
TRAINING_CLASS, "class1",
TRAINING_CLASS, "class2"
));
//class2
addDoc(writer, buildLuceneDocument(ID, "5",
TITLE, "word4 word4 word4",
CONTENT, "word5 word5",
AUTHOR, "Name1 Surname1",
TRAINING_CLASS, "class6",
TRAINING_CLASS, "class4"
));
addDoc(writer, buildLuceneDocument(ID, "6",
TITLE, "word4 word4",
CONTENT, "word5",
AUTHOR, "Name1 Surname1",
TRAINING_CLASS, "class5",
TRAINING_CLASS, "class4"
));
addDoc(writer, buildLuceneDocument(ID, "7",
TITLE, "word4 word4 word4",
CONTENT, "word5 word5 word5",
AUTHOR, "Name1 Surname1",
TRAINING_CLASS, "class6",
TRAINING_CLASS, "class4"
));
addDoc(writer, buildLuceneDocument(ID, "8",
TITLE, "word4",
CONTENT, "word5 word5 word5 word5",
AUTHOR, "Name1 Surname1",
TRAINING_CLASS, "class6",
TRAINING_CLASS, "class4"
));
reader = writer.getReader();
writer.close();
searcher = newSearcher(reader);
}
public static Document buildLuceneDocument(Object... fieldsAndValues) {
Document luceneDoc = new Document();
for (int i=0; i<fieldsAndValues.length; i+=2) {
luceneDoc.add(newTextField((String)fieldsAndValues[i], (String)fieldsAndValues[i+1], Field.Store.YES));
}
return luceneDoc;
}
private int addDoc(RandomIndexWriter writer, Document doc) throws IOException {
writer.addDocument(doc);
return writer.numDocs() - 1;
}
}