Menu
▾
▴
dl-learner-svn
|
From: <jen...@us...> - 2010-09-21 12:38:52
|
Revision: 2327
http://dl-learner.svn.sourceforge.net/dl-learner/?rev=2327&view=rev
Author: jenslehmann
Date: 2010-09-21 12:38:44 +0000 (Tue, 21 Sep 2010)
Log Message:
-----------
fixed remaining compile errors
Modified Paths:
--------------
trunk/scripts/src/main/java/org/dllearner/scripts/CrossValidation.java
trunk/scripts/src/main/java/org/dllearner/scripts/NestedCrossValidation.java
trunk/scripts/src/main/java/org/dllearner/scripts/TestValidation.java
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/CrossValidation.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/CrossValidation.java 2010-09-21 12:28:13 UTC (rev 2326)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/CrossValidation.java 2010-09-21 12:38:44 UTC (rev 2327)
@@ -118,7 +118,7 @@
} catch (FileNotFoundException e) {
// TODO Auto-generated catch block
e.printStackTrace();
- } catch (ParseException e) {
+ } catch (org.dllearner.confparser.ParseException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
@@ -208,7 +208,7 @@
} catch (FileNotFoundException e) {
// TODO Auto-generated catch block
e.printStackTrace();
- } catch (ParseException e) {
+ } catch (org.dllearner.confparser.ParseException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/NestedCrossValidation.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/NestedCrossValidation.java 2010-09-21 12:28:13 UTC (rev 2326)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/NestedCrossValidation.java 2010-09-21 12:38:44 UTC (rev 2327)
@@ -102,8 +102,9 @@
* @throws IOException
* @throws ParseException
* @throws ComponentInitException
+ * @throws org.dllearner.confparser.ParseException
*/
- public static void main(String[] args) throws IOException, ComponentInitException, ParseException {
+ public static void main(String[] args) throws IOException, ComponentInitException, ParseException, org.dllearner.confparser.ParseException {
OptionParser parser = new OptionParser();
parser.acceptsAll(asList("h", "?", "help"), "Show help.");
@@ -162,7 +163,7 @@
}
- public NestedCrossValidation(File confFile, int outerFolds, int innerFolds, String parameter, int startValue, int endValue, boolean verbose) throws FileNotFoundException, ComponentInitException, ParseException {
+ public NestedCrossValidation(File confFile, int outerFolds, int innerFolds, String parameter, int startValue, int endValue, boolean verbose) throws FileNotFoundException, ComponentInitException, ParseException, org.dllearner.confparser.ParseException {
DecimalFormat df = new DecimalFormat();
ComponentManager cm = ComponentManager.getInstance();
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/TestValidation.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/TestValidation.java 2010-09-21 12:28:13 UTC (rev 2326)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/TestValidation.java 2010-09-21 12:38:44 UTC (rev 2327)
@@ -44,7 +44,7 @@
private static Logger logger = Logger.getRootLogger();
- public static void main(String args[]) throws ComponentInitException, FileNotFoundException, ParseException {
+ public static void main(String args[]) throws ComponentInitException, FileNotFoundException, ParseException, org.dllearner.confparser.ParseException {
// create logger (a simple logger which outputs
// its messages to the console)
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <jen...@us...> - 2011-08-03 13:00:25
|
Revision: 2978
http://dl-learner.svn.sourceforge.net/dl-learner/?rev=2978&view=rev
Author: jenslehmann
Date: 2011-08-03 13:00:19 +0000 (Wed, 03 Aug 2011)
Log Message:
-----------
started enrichment evaluator
Modified Paths:
--------------
trunk/scripts/src/main/java/org/dllearner/scripts/matching/GeneralMatcher.java
Added Paths:
-----------
trunk/scripts/src/main/java/org/dllearner/scripts/evaluation/EnrichmentEvaluation.java
Added: trunk/scripts/src/main/java/org/dllearner/scripts/evaluation/EnrichmentEvaluation.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/evaluation/EnrichmentEvaluation.java (rev 0)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/evaluation/EnrichmentEvaluation.java 2011-08-03 13:00:19 UTC (rev 2978)
@@ -0,0 +1,97 @@
+/**
+ * Copyright (C) 2007-2008, Jens Lehmann
+ *
+ * This file is part of DL-Learner.
+ *
+ * DL-Learner is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * DL-Learner is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ *
+ */
+package org.dllearner.scripts.evaluation;
+
+import java.util.Set;
+import java.util.TreeSet;
+
+import org.aksw.commons.sparql.sparqler.SelectPaginated;
+import org.aksw.commons.sparql.sparqler.Sparqler;
+import org.aksw.commons.sparql.sparqler.SparqlerHttp;
+import org.dllearner.algorithms.properties.SubPropertyOfAxiomLearner;
+import org.dllearner.core.owl.ObjectProperty;
+import org.dllearner.kb.SparqlEndpointKS;
+import org.dllearner.kb.sparql.SparqlEndpoint;
+
+import com.hp.hpl.jena.query.QuerySolution;
+
+/**
+ * Evaluation of enrichment algorithms on DBpedia (Live).
+ *
+ * @author Jens Lehmann
+ *
+ */
+public class EnrichmentEvaluation {
+
+ // max. execution time for each learner for each entity
+ private int maxExecutionTimeInSeconds = 10;
+
+ public EnrichmentEvaluation() {
+
+ }
+
+ public void start() {
+
+ // create DBpedia Live knowledge source
+ SparqlEndpoint se = SparqlEndpoint.getEndpointDBpediaLiveAKSW();
+
+ Set<ObjectProperty> properties = getAllObjectProperties(se);
+ System.out.println(properties);
+
+ SparqlEndpointKS ks = new SparqlEndpointKS(se);
+
+ SubPropertyOfAxiomLearner learner = new SubPropertyOfAxiomLearner(ks);
+ learner.setMaxExecutionTimeInSeconds(10);
+
+ }
+
+ private void getAllClasses() {
+
+
+
+ }
+
+ private Set<ObjectProperty> getAllObjectProperties(SparqlEndpoint se) {
+ Set<ObjectProperty> properties = new TreeSet<ObjectProperty>();
+ Sparqler x = new SparqlerHttp(se.getURL().toString());
+ SelectPaginated q = new SelectPaginated(x, "PREFIX owl: <http://www.w3.org/2002/07/owl#> SELECT ?p WHERE {?p a owl:ObjectProperty}", 1000);
+ while(q.hasNext()) {
+ QuerySolution qs = q.next();
+ System.out.println(qs);
+ properties.add(new ObjectProperty(qs.getResource("p").getURI()));
+ }
+ return properties;
+ }
+
+ public void printResultsPlain() {
+
+ }
+
+ public void printResultsLaTeX() {
+
+ }
+
+ public static void main(String[] args) {
+ EnrichmentEvaluation ee = new EnrichmentEvaluation();
+ ee.start();
+ ee.printResultsPlain();
+ }
+
+}
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/matching/GeneralMatcher.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/matching/GeneralMatcher.java 2011-08-03 05:58:49 UTC (rev 2977)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/matching/GeneralMatcher.java 2011-08-03 13:00:19 UTC (rev 2978)
@@ -24,6 +24,9 @@
import java.util.LinkedList;
import java.util.Set;
+import org.aksw.commons.sparql.sparqler.SelectPaginated;
+import org.aksw.commons.sparql.sparqler.Sparqler;
+import org.aksw.commons.sparql.sparqler.SparqlerHttp;
import org.dllearner.algorithm.qtl.QTL;
import org.dllearner.core.ComponentInitException;
import org.dllearner.core.LearningProblemUnsupportedException;
@@ -31,6 +34,8 @@
import org.dllearner.kb.sparql.SparqlEndpoint;
import org.dllearner.learningproblems.PosOnlyLP;
+import com.hp.hpl.jena.query.QuerySolution;
+
/**
*
*
@@ -45,6 +50,13 @@
// option 1: read links from file
// option 2: use SPARQL to get links
+ // use AKSW commons SPARQL API
+ Sparqler x = new SparqlerHttp(sparqlEndpoint1.getURL().toString());
+ SelectPaginated q = new SelectPaginated(x, "query", 1000);
+ while(q.hasNext()) {
+ QuerySolution qs = q.next();
+ }
+
// phase 2: create learning problems
// phase 3: execute learning algorithms
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <lor...@us...> - 2011-08-04 16:05:05
|
Revision: 3001
http://dl-learner.svn.sourceforge.net/dl-learner/?rev=3001&view=rev
Author: lorenz_b
Date: 2011-08-04 16:04:59 +0000 (Thu, 04 Aug 2011)
Log Message:
-----------
Small changes to compile successfully.
Modified Paths:
--------------
trunk/scripts/src/main/java/org/dllearner/scripts/improveWikipedia/DBpediaClassLearnerCELOE.java
trunk/scripts/src/main/java/org/dllearner/scripts/matching/GeneralMatcher.java
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/improveWikipedia/DBpediaClassLearnerCELOE.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/improveWikipedia/DBpediaClassLearnerCELOE.java 2011-08-04 14:24:53 UTC (rev 3000)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/improveWikipedia/DBpediaClassLearnerCELOE.java 2011-08-04 16:04:59 UTC (rev 3001)
@@ -19,6 +19,17 @@
*/
package org.dllearner.scripts.improveWikipedia;
+import java.io.File;
+import java.io.IOException;
+import java.net.URL;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.Map;
+import java.util.Set;
+import java.util.SortedSet;
+import java.util.TreeSet;
+
import org.aksw.commons.sparql.core.ResultSetRenderer;
import org.aksw.commons.sparql.core.SparqlEndpoint;
import org.aksw.commons.sparql.core.SparqlTemplate;
@@ -26,12 +37,17 @@
import org.aksw.commons.sparql.core.impl.HttpSparqlEndpoint;
import org.apache.velocity.VelocityContext;
import org.dllearner.algorithms.celoe.CELOE;
+import org.dllearner.core.AbstractReasonerComponent;
import org.dllearner.core.ComponentManager;
import org.dllearner.core.LearningProblemUnsupportedException;
import org.dllearner.core.OntologyFormat;
-import org.dllearner.core.AbstractReasonerComponent;
import org.dllearner.core.configurators.CELOEConfigurator;
-import org.dllearner.core.owl.*;
+import org.dllearner.core.owl.Description;
+import org.dllearner.core.owl.EquivalentClassesAxiom;
+import org.dllearner.core.owl.Individual;
+import org.dllearner.core.owl.KB;
+import org.dllearner.core.owl.NamedClass;
+import org.dllearner.core.owl.Thing;
import org.dllearner.gui.Config;
import org.dllearner.gui.ConfigSave;
import org.dllearner.kb.sparql.SparqlKnowledgeSource;
@@ -42,11 +58,6 @@
import org.dllearner.utilities.datastructures.SetManipulation;
import org.dllearner.utilities.datastructures.SortedSetTuple;
-import java.io.File;
-import java.io.IOException;
-import java.net.URL;
-import java.util.*;
-
/**
* A script, which learns definitions / super classes of classes in the DBpedia ontology.
*
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/matching/GeneralMatcher.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/matching/GeneralMatcher.java 2011-08-04 14:24:53 UTC (rev 3000)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/matching/GeneralMatcher.java 2011-08-04 16:04:59 UTC (rev 3001)
@@ -24,9 +24,6 @@
import java.util.LinkedList;
import java.util.Set;
-import org.aksw.commons.sparql.sparqler.SelectPaginated;
-import org.aksw.commons.sparql.sparqler.Sparqler;
-import org.aksw.commons.sparql.sparqler.SparqlerHttp;
import org.dllearner.algorithm.qtl.QTL;
import org.dllearner.core.ComponentInitException;
import org.dllearner.core.LearningProblemUnsupportedException;
@@ -34,8 +31,6 @@
import org.dllearner.kb.sparql.SparqlEndpoint;
import org.dllearner.learningproblems.PosOnlyLP;
-import com.hp.hpl.jena.query.QuerySolution;
-
/**
*
*
@@ -51,11 +46,11 @@
// option 2: use SPARQL to get links
// use AKSW commons SPARQL API
- Sparqler x = new SparqlerHttp(sparqlEndpoint1.getURL().toString());
- SelectPaginated q = new SelectPaginated(x, "query", 1000);
- while(q.hasNext()) {
- QuerySolution qs = q.next();
- }
+// Sparqler x = new SparqlerHttp(sparqlEndpoint1.getURL().toString());
+// SelectPaginated q = new SelectPaginated(x, "query", 1000);
+// while(q.hasNext()) {
+// QuerySolution qs = q.next();
+// }
// phase 2: create learning problems
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <lor...@us...> - 2012-02-01 11:36:56
|
Revision: 3561
http://dl-learner.svn.sourceforge.net/dl-learner/?rev=3561&view=rev
Author: lorenz_b
Date: 2012-02-01 11:36:50 +0000 (Wed, 01 Feb 2012)
Log Message:
-----------
Added renderer for justifications.
Modified Paths:
--------------
trunk/scripts/src/main/java/org/dllearner/scripts/ApproximateDebugging.java
trunk/scripts/src/main/java/org/dllearner/scripts/SPARQLSampleDebugging.java
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/ApproximateDebugging.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/ApproximateDebugging.java 2012-02-01 11:35:39 UTC (rev 3560)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/ApproximateDebugging.java 2012-02-01 11:36:50 UTC (rev 3561)
@@ -1,30 +1,34 @@
package org.dllearner.scripts;
+import java.io.BufferedInputStream;
import java.io.ByteArrayInputStream;
import java.io.ByteArrayOutputStream;
import java.io.File;
+import java.io.FileInputStream;
import java.io.IOException;
+import java.io.InputStream;
import java.io.PrintWriter;
import java.net.URL;
import java.util.Collections;
-import java.util.Comparator;
+import java.util.HashMap;
import java.util.HashSet;
+import java.util.Map;
+import java.util.Map.Entry;
import java.util.Set;
import java.util.SortedSet;
import java.util.TreeSet;
+import org.apache.commons.compress.compressors.CompressorInputStream;
+import org.apache.commons.compress.compressors.CompressorStreamFactory;
import org.apache.log4j.ConsoleAppender;
import org.apache.log4j.FileAppender;
import org.apache.log4j.Level;
import org.apache.log4j.Logger;
import org.apache.log4j.SimpleLayout;
-import org.dllearner.algorithm.qtl.util.ModelGenerator;
-import org.dllearner.algorithm.qtl.util.ModelGenerator.Strategy;
import org.dllearner.kb.sparql.ConciseBoundedDescriptionGenerator;
import org.dllearner.kb.sparql.ConciseBoundedDescriptionGeneratorImpl;
import org.dllearner.kb.sparql.ExtractionDBCache;
import org.dllearner.kb.sparql.SparqlEndpoint;
-import org.dllearner.kb.sparql.SymmetricConciseBoundedDescriptionGeneratorImpl;
import org.mindswap.pellet.PelletOptions;
import org.semanticweb.owlapi.apibinding.OWLManager;
import org.semanticweb.owlapi.io.RDFXMLOntologyFormat;
@@ -35,7 +39,6 @@
import org.semanticweb.owlapi.model.OWLDataFactory;
import org.semanticweb.owlapi.model.OWLDisjointClassesAxiom;
import org.semanticweb.owlapi.model.OWLEntity;
-import org.semanticweb.owlapi.model.OWLException;
import org.semanticweb.owlapi.model.OWLIndividual;
import org.semanticweb.owlapi.model.OWLObjectProperty;
import org.semanticweb.owlapi.model.OWLObjectPropertyAssertionAxiom;
@@ -48,7 +51,6 @@
import uk.ac.manchester.cs.owl.owlapi.OWLDataFactoryImpl;
import uk.ac.manchester.cs.owlapi.modularity.ModuleType;
-import weka.attributeSelection.ConsistencySubsetEval;
import com.clarkparsia.modularity.ModularityUtils;
import com.clarkparsia.owlapi.explanation.PelletExplanation;
@@ -84,6 +86,12 @@
public ApproximateDebugging(OWLOntology schema, OWLOntology data) throws OWLOntologyCreationException {
this.schema = schema;
this.data = data;
+ System.out.println(data.getLogicalAxiomCount());
+ for(OWLAxiom ax : new HashSet<OWLAxiom>(data.getLogicalAxioms())){
+ if(!ax.getDataPropertiesInSignature().isEmpty()){
+ man.removeAxiom(data, ax);
+ }
+ }System.out.println(data.getLogicalAxiomCount());
Set<OWLOntology> ontologies = new HashSet<OWLOntology>();
ontologies.add(schema);
@@ -181,7 +189,8 @@
logger.info("done in " + (System.currentTimeMillis()-startTime) + "ms.");
logger.info("Module size: " + module.size());
reasoner = PelletReasonerFactory.getInstance().createNonBufferingReasoner(OntologyUtils.getOntologyFromAxioms(module));
- expGen = new PelletExplanation(reasoner);
+// expGen = new PelletExplanation(reasoner);
+ expGen = new PelletExplanation(schema, false);
logger.info("Computing explanations...");
startTime = System.currentTimeMillis();
Set<Set<OWLAxiom>> temp = expGen.getUnsatisfiableExplanations(unsatClass, 10);
@@ -231,9 +240,9 @@
}
public Set<Set<OWLAxiom>> computeExplanationsDefault(int limit){
- reasoner = PelletReasonerFactory.getInstance().createNonBufferingReasoner(schema);
+ reasoner = PelletReasonerFactory.getInstance().createNonBufferingReasoner(ontology);
reasoner.isConsistent();
- logger.info("Computing inconsistency explanations with only Pellet...");
+ logger.info("Computing inconsistency explanations with Pellet only...");
long startTime = System.currentTimeMillis();
PelletExplanation expGen = new PelletExplanation(reasoner);
Set<Set<OWLAxiom>> explanations = expGen.getInconsistencyExplanations(limit);
@@ -312,7 +321,7 @@
for(OWLObjectProperty prop : extractObjectProperties(AxiomType.FUNCTIONAL_OBJECT_PROPERTY)){
OWLAxiom axiom = factory.getOWLFunctionalObjectPropertyAxiom(prop);
- String queryString = "SELECT * WHERE {?s <%s> ?o1. ?s <%s> ?o2. FILTER(?o1 != ?o2)}".replace("%s", prop.toStringID());
+ String queryString = "SELECT DISTINCT * WHERE {?s <%s> ?o1. ?s <%s> ?o2. FILTER(?o1 != ?o2)}".replace("%s", prop.toStringID());
Query query = QueryFactory.create(queryString) ;
QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
try {
@@ -427,6 +436,7 @@
for(OWLAxiom ax : ontology.getAxioms(axiomType)){
properties.add(((OWLObjectPropertyCharacteristicAxiom)ax).getProperty().asOWLObjectProperty());
}
+// properties.retainAll(data.getObjectPropertiesInSignature());
return properties;
}
@@ -472,6 +482,11 @@
* @throws Exception
*/
public static void main(String[] args) throws Exception {
+ if(args.length != 1){
+ System.out.println("Usage: ApproximateDebugging <Schema-Ontology>");
+ System.exit(0);
+ }
+
Logger.getRootLogger().setLevel(Level.INFO);
Logger.getRootLogger().removeAllAppenders();
Logger.getRootLogger().addAppender(new ConsoleAppender(new SimpleLayout()));
@@ -487,14 +502,55 @@
Model model = cbdGen.getConciseBoundedDescription(resource, 3);
OWLOntology data = convert(model);
+ InputStream is = new BufferedInputStream(new FileInputStream(args[0]));
+ CompressorInputStream in = new CompressorStreamFactory().createCompressorInputStream("bzip2", is);
+
OWLOntologyManager man = OWLManager.createOWLOntologyManager();
- OWLOntology schema = man.loadOntologyFromOntologyDocument(new File("src/main/resources/dbpedia_0.75_no_datapropaxioms.owl"));
+ OWLOntology schema = man.loadOntologyFromOntologyDocument(in);
ApproximateDebugging debug = new ApproximateDebugging(schema, data);
+
+ Set<Set<OWLAxiom>> explanations2 = debug.computeExplanationsDefault(3);
+ ManchesterSyntaxExplanationRenderer renderer = new ManchesterSyntaxExplanationRenderer();
+ PrintWriter out = new PrintWriter( System.out );
+ renderer.startRendering( out );
+ for(Set<OWLAxiom> exp : explanations2){
+ out.flush();
+ renderer.render( Collections.singleton(exp) );
+ }
+ renderer.endRendering();
+ for(Entry<AxiomType<? extends OWLAxiom>, Integer> entry : getAxiomTypeCount(explanations2).entrySet()){
+ System.out.println(entry.getKey() + ": " + entry.getValue());
+ }
+
Set<Set<OWLAxiom>> explanations1 = debug.computeInconsistencyExplanations();
- Set<Set<OWLAxiom>> explanations2 = debug.computeExplanationsDefault(500);
- System.out.println(explanations1.size());
- System.out.println(explanations2.size());
+ for(Set<OWLAxiom> exp : explanations1){
+ System.out.println(exp);
+ }
+
+ for(Entry<AxiomType<? extends OWLAxiom>, Integer> entry : getAxiomTypeCount(explanations1).entrySet()){
+ System.out.println(entry.getKey() + ": " + entry.getValue());
+ }
+//
+// System.out.println(explanations1.size());
+// System.out.println(explanations2.size());
}
+
+ private static Map<AxiomType<? extends OWLAxiom>, Integer> getAxiomTypeCount(Set<Set<OWLAxiom>> explanations){
+ Map<AxiomType<? extends OWLAxiom>, Integer> axiomType2CountMap = new HashMap<AxiomType<? extends OWLAxiom>, Integer>();
+
+ for(Set<OWLAxiom> exp : explanations){
+ for(OWLAxiom ax : exp){
+ Integer cnt = axiomType2CountMap.get(ax.getAxiomType());
+ if(cnt == null){
+ cnt = Integer.valueOf(0);
+ }
+ cnt = Integer.valueOf(cnt + 1);
+ axiomType2CountMap.put(ax.getAxiomType(), cnt);
+ }
+ }
+
+ return axiomType2CountMap;
+ }
}
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/SPARQLSampleDebugging.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/SPARQLSampleDebugging.java 2012-02-01 11:35:39 UTC (rev 3560)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/SPARQLSampleDebugging.java 2012-02-01 11:36:50 UTC (rev 3561)
@@ -6,6 +6,7 @@
import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.FileReader;
+import java.io.FileWriter;
import java.io.IOException;
import java.io.InputStream;
import java.io.PrintWriter;
@@ -29,7 +30,6 @@
import java.util.TreeSet;
import java.util.prefs.Preferences;
-import org.apache.commons.compress.compressors.CompressorInputStream;
import org.apache.commons.compress.compressors.CompressorStreamFactory;
import org.apache.log4j.ConsoleAppender;
import org.apache.log4j.FileAppender;
@@ -54,6 +54,7 @@
import org.semanticweb.owlapi.model.OWLClass;
import org.semanticweb.owlapi.model.OWLDataFactory;
import org.semanticweb.owlapi.model.OWLDataProperty;
+import org.semanticweb.owlapi.model.OWLException;
import org.semanticweb.owlapi.model.OWLFunctionalDataPropertyAxiom;
import org.semanticweb.owlapi.model.OWLFunctionalObjectPropertyAxiom;
import org.semanticweb.owlapi.model.OWLIndividual;
@@ -473,16 +474,32 @@
logger.info(exp + "\n");
}
Map<AxiomType, Integer> axiomType2CountMap = new HashMap<AxiomType, Integer>();
- for(Set<OWLAxiom> explanation : explanations){
- logger.info(explanation);
- for(OWLAxiom axiom : explanation){
- Integer cnt = axiomType2CountMap.get(axiom.getAxiomType());
- if(cnt == null){
- cnt = Integer.valueOf(0);
+ ManchesterSyntaxExplanationRenderer renderer = new ManchesterSyntaxExplanationRenderer();
+ try {
+ FileWriter out = new FileWriter( "log/justifications.txt" );
+ renderer.startRendering(out );
+ for(Set<OWLAxiom> explanation : explanations){
+ logger.info(explanation);
+ out.flush();
+ try {
+ renderer.render( Collections.singleton(explanation) );
+ } catch (UnsupportedOperationException e) {
+ e.printStackTrace();
+ } catch (OWLException e) {
+ e.printStackTrace();
}
- cnt = Integer.valueOf(cnt + 1);
- axiomType2CountMap.put(axiom.getAxiomType(), cnt);
+ for(OWLAxiom axiom : explanation){
+ Integer cnt = axiomType2CountMap.get(axiom.getAxiomType());
+ if(cnt == null){
+ cnt = Integer.valueOf(0);
+ }
+ cnt = Integer.valueOf(cnt + 1);
+ axiomType2CountMap.put(axiom.getAxiomType(), cnt);
+ }
}
+ renderer.endRendering();
+ } catch (IOException e) {
+ e.printStackTrace();
}
logger.info("Axiom type count:");
for(Entry<AxiomType, Integer> entry : axiomType2CountMap.entrySet()){
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <lor...@us...> - 2013-01-30 09:55:31
|
Revision: 3893
http://dl-learner.svn.sourceforge.net/dl-learner/?rev=3893&view=rev
Author: lorenz_b
Date: 2013-01-30 09:55:24 +0000 (Wed, 30 Jan 2013)
Log Message:
-----------
Continued matching.
Modified Paths:
--------------
trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatching.java
Added Paths:
-----------
trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatchingTest.java
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatching.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatching.java 2013-01-25 13:04:28 UTC (rev 3892)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatching.java 2013-01-30 09:55:24 UTC (rev 3893)
@@ -1,6 +1,5 @@
package org.dllearner.scripts;
-import java.net.URL;
import java.text.DecimalFormat;
import java.util.HashMap;
import java.util.List;
@@ -10,10 +9,7 @@
import java.util.SortedSet;
import java.util.TreeSet;
-import org.apache.log4j.ConsoleAppender;
-import org.apache.log4j.Level;
import org.apache.log4j.Logger;
-import org.apache.log4j.PatternLayout;
import org.dllearner.algorithms.celoe.CELOE;
import org.dllearner.core.AbstractLearningProblem;
import org.dllearner.core.AbstractReasonerComponent;
@@ -24,7 +20,6 @@
import org.dllearner.core.owl.NamedClass;
import org.dllearner.core.owl.ObjectProperty;
import org.dllearner.kb.SparqlEndpointKS;
-import org.dllearner.kb.extraction.ExtractionAlgorithm;
import org.dllearner.kb.sparql.ExtractionDBCache;
import org.dllearner.kb.sparql.SPARQLTasks;
import org.dllearner.kb.sparql.SparqlEndpoint;
@@ -35,14 +30,11 @@
import org.dllearner.reasoning.FastInstanceChecker;
import org.dllearner.reasoning.SPARQLReasoner;
import org.dllearner.utilities.datastructures.Datastructures;
+import org.dllearner.utilities.datastructures.SetManipulation;
import org.dllearner.utilities.datastructures.SortedSetTuple;
import org.dllearner.utilities.examples.AutomaticNegativeExampleFinderSPARQL2;
-import org.dllearner.utilities.owl.DLLearnerDescriptionConvertVisitor;
import org.dllearner.utilities.owl.OWLAPIDescriptionConvertVisitor;
-import org.semanticweb.owlapi.io.ToStringRenderer;
-import uk.ac.manchester.cs.owl.owlapi.mansyntaxrenderer.ManchesterOWLSyntaxOWLObjectRendererImpl;
-
import com.hp.hpl.jena.query.Query;
import com.hp.hpl.jena.query.QueryFactory;
import com.hp.hpl.jena.query.QuerySolution;
@@ -132,9 +124,10 @@
logger.info(nc);
// get all via owl:sameAs related individuals
SortedSet<Individual> individuals = getRelatedIndividualsNamespaceAware(source, nc, target.getNamespace());
- logger.info(individuals);
+ logger.info(individuals.size());
//learn concept in KB2 based on the examples
if(individuals.size() >= 3){
+ individuals = SetManipulation.fuzzyShrinkInd(individuals, 10);
List<? extends EvaluatedDescription> learnedClassExpressions = learnClassExpression(target, individuals);
mapping.put(nc, learnedClassExpressions);
}
@@ -260,6 +253,9 @@
String uri = object.asResource().getURI();
//workaround for World Factbook - should be removed later
uri = uri.replace("http://www4.wiwiss.fu-berlin.de/factbook/resource/", "http://wifo5-03.informatik.uni-mannheim.de/factbook/resource/");
+ //workaround for OpenCyc - should be removed later
+ uri = uri.replace("http://sw.cyc.com", "http://sw.opencyc.org");
+
relatedIndividuals.add(new Individual(uri));
}
}
@@ -329,40 +325,5 @@
return cache;
}
-
-
}
-
- /**
- * @param args
- */
- public static void main(String[] args) throws Exception{
- //render output
- ToStringRenderer.getInstance().setRenderer(new ManchesterOWLSyntaxOWLObjectRendererImpl());
- //set logging properties
- Logger.getLogger(SparqlKnowledgeSource.class).setLevel(Level.WARN);
- Logger.getLogger(ExtractionAlgorithm.class).setLevel(Level.WARN);
- Logger.getLogger(org.dllearner.kb.extraction.Manager.class).setLevel(Level.WARN);
- Logger.getRootLogger().removeAllAppenders();
- Logger.getRootLogger().addAppender(new ConsoleAppender(new PatternLayout("%m%n")));
- // KB2
- SparqlEndpoint endpoint1 = SparqlEndpoint.getEndpointDBpedia();
- ExtractionDBCache cache1 = new ExtractionDBCache("cache");
- String namespace1 = "http://dbpedia.org/resource/";
- KnowledgeBase kb1 = new KnowledgeBase(endpoint1, cache1, namespace1);
- // KB2
- SparqlEndpoint endpoint2 = new SparqlEndpoint(new URL("http://wifo5-03.informatik.uni-mannheim.de/factbook/sparql"));
- ExtractionDBCache cache2 = new ExtractionDBCache("cache");
- //TODO problem with World Factbook is that old FU Berlin server is useless because of bugs and current version
- //is provide by University Of Mannheim now with another namespace http://wifo5-03.informatik.uni-mannheim.de/factbook/resource/
- //but the DBpedia links are still to the old D2R server instance
- //workaround: replace namespace before learning
- String namespace2 = "http://www4.wiwiss.fu-berlin.de/factbook/resource/";
- KnowledgeBase kb2 = new KnowledgeBase(endpoint2, cache2, namespace2);
-
- OntologyMatching matcher = new OntologyMatching(kb1, kb2);
- matcher.start();
-
- }
-
}
Added: trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatchingTest.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatchingTest.java (rev 0)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatchingTest.java 2013-01-30 09:55:24 UTC (rev 3893)
@@ -0,0 +1,85 @@
+package org.dllearner.scripts;
+
+import java.net.URL;
+
+import org.apache.log4j.ConsoleAppender;
+import org.apache.log4j.Level;
+import org.apache.log4j.Logger;
+import org.apache.log4j.PatternLayout;
+import org.dllearner.kb.extraction.ExtractionAlgorithm;
+import org.dllearner.kb.sparql.ExtractionDBCache;
+import org.dllearner.kb.sparql.SparqlEndpoint;
+import org.dllearner.kb.sparql.SparqlKnowledgeSource;
+import org.dllearner.scripts.OntologyMatching.KnowledgeBase;
+import org.junit.Before;
+import org.junit.Test;
+import org.semanticweb.owlapi.io.ToStringRenderer;
+
+import uk.ac.manchester.cs.owl.owlapi.mansyntaxrenderer.ManchesterOWLSyntaxOWLObjectRendererImpl;
+
+public class OntologyMatchingTest {
+
+ private KnowledgeBase dbpedia;
+ private KnowledgeBase worldFactBook;
+ private KnowledgeBase openCyc;
+ private KnowledgeBase linkedGeoData;
+
+ @Before
+ public void setUp() throws Exception {
+ // render output
+ ToStringRenderer.getInstance().setRenderer(new ManchesterOWLSyntaxOWLObjectRendererImpl());
+ // set logging properties
+ Logger.getLogger(SparqlKnowledgeSource.class).setLevel(Level.WARN);
+ Logger.getLogger(ExtractionAlgorithm.class).setLevel(Level.WARN);
+ Logger.getLogger(org.dllearner.kb.extraction.Manager.class).setLevel(Level.WARN);
+ Logger.getRootLogger().removeAllAppenders();
+ Logger.getRootLogger().addAppender(new ConsoleAppender(new PatternLayout("%m%n")));
+
+ //DBpedia
+ SparqlEndpoint endpoint = SparqlEndpoint.getEndpointDBpedia();
+ ExtractionDBCache cache = new ExtractionDBCache("cache");
+ String namespace = "http://dbpedia.org/resource/";
+ dbpedia = new KnowledgeBase(endpoint, cache, namespace);
+
+ //World Factbook
+ //TODO problem with World Factbook is that old FU Berlin server is useless because of bugs and current version
+ //is provide by University Of Mannheim now with another namespace http://wifo5-03.informatik.uni-mannheim.de/factbook/resource/
+ //but the DBpedia links are still to the old D2R server instance
+ //workaround: replace namespace before learning
+ endpoint = new SparqlEndpoint(new URL("http://wifo5-03.informatik.uni-mannheim.de/factbook/sparql"));
+ cache = new ExtractionDBCache("cache");
+ namespace = "http://www4.wiwiss.fu-berlin.de/factbook/resource/";
+ worldFactBook = new KnowledgeBase(endpoint, cache, namespace);
+
+ //local OpenCyc
+ endpoint = new SparqlEndpoint(new URL("http://localhost:8890/sparql"));
+ cache = new ExtractionDBCache("cache");
+ namespace = "http://sw.cyc.com";
+ openCyc = new KnowledgeBase(endpoint, cache, namespace);
+
+ //LinkedGeoData
+ endpoint = new SparqlEndpoint(new URL("http://linkedgeodata.org/sparql"));
+ cache = new ExtractionDBCache("cache");
+ namespace = "http://linkedgeodata.org/triplify/";
+ linkedGeoData = new KnowledgeBase(endpoint, cache, namespace);
+ }
+
+ @Test
+ public void testDBpediaWorldFactbook() {
+ OntologyMatching matcher = new OntologyMatching(dbpedia, worldFactBook);
+ matcher.start();
+ }
+
+ @Test
+ public void testDBpediaOpenCyc() {
+ OntologyMatching matcher = new OntologyMatching(dbpedia, openCyc);
+ matcher.start();
+ }
+
+ @Test
+ public void testDBpediaLinkedGeoData() {
+ OntologyMatching matcher = new OntologyMatching(dbpedia, linkedGeoData);
+ matcher.start();
+ }
+
+}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
|
From: <lor...@us...> - 2013-02-06 18:01:57
|
Revision: 3895
http://dl-learner.svn.sourceforge.net/dl-learner/?rev=3895&view=rev
Author: lorenz_b
Date: 2013-02-06 18:01:42 +0000 (Wed, 06 Feb 2013)
Log Message:
-----------
Continued matching component.
Modified Paths:
--------------
trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatching.java
trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatchingTest.java
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatching.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatching.java 2013-02-06 17:59:21 UTC (rev 3894)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatching.java 2013-02-06 18:01:42 UTC (rev 3895)
@@ -1,7 +1,15 @@
package org.dllearner.scripts;
+import java.io.ByteArrayInputStream;
+import java.io.ByteArrayOutputStream;
+import java.io.File;
+import java.io.FileNotFoundException;
+import java.io.FileOutputStream;
import java.text.DecimalFormat;
+import java.util.ArrayList;
import java.util.HashMap;
+import java.util.HashSet;
+import java.util.Iterator;
import java.util.List;
import java.util.Map;
import java.util.Map.Entry;
@@ -15,11 +23,15 @@
import org.dllearner.core.AbstractReasonerComponent;
import org.dllearner.core.ComponentInitException;
import org.dllearner.core.EvaluatedDescription;
+import org.dllearner.core.KnowledgeSource;
import org.dllearner.core.owl.Description;
import org.dllearner.core.owl.Individual;
import org.dllearner.core.owl.NamedClass;
import org.dllearner.core.owl.ObjectProperty;
+import org.dllearner.kb.OWLAPIOntology;
import org.dllearner.kb.SparqlEndpointKS;
+import org.dllearner.kb.sparql.ConciseBoundedDescriptionGenerator;
+import org.dllearner.kb.sparql.ConciseBoundedDescriptionGeneratorImpl;
import org.dllearner.kb.sparql.ExtractionDBCache;
import org.dllearner.kb.sparql.SPARQLTasks;
import org.dllearner.kb.sparql.SparqlEndpoint;
@@ -29,18 +41,33 @@
import org.dllearner.learningproblems.PosOnlyLP;
import org.dllearner.reasoning.FastInstanceChecker;
import org.dllearner.reasoning.SPARQLReasoner;
+import org.dllearner.utilities.LabelShortFormProvider;
import org.dllearner.utilities.datastructures.Datastructures;
import org.dllearner.utilities.datastructures.SetManipulation;
import org.dllearner.utilities.datastructures.SortedSetTuple;
import org.dllearner.utilities.examples.AutomaticNegativeExampleFinderSPARQL2;
import org.dllearner.utilities.owl.OWLAPIDescriptionConvertVisitor;
+import org.semanticweb.owlapi.apibinding.OWLManager;
+import org.semanticweb.owlapi.io.UnparsableOntologyException;
+import org.semanticweb.owlapi.model.OWLOntology;
+import org.semanticweb.owlapi.model.OWLOntologyCreationException;
+import org.semanticweb.owlapi.model.OWLOntologyManager;
+import uk.ac.manchester.cs.owl.owlapi.mansyntaxrenderer.ManchesterOWLSyntaxOWLObjectRendererImpl;
+
+import com.google.common.collect.Sets;
+import com.hp.hpl.jena.ontology.OntModel;
+import com.hp.hpl.jena.query.ParameterizedSparqlString;
import com.hp.hpl.jena.query.Query;
+import com.hp.hpl.jena.query.QueryExecutionFactory;
import com.hp.hpl.jena.query.QueryFactory;
import com.hp.hpl.jena.query.QuerySolution;
import com.hp.hpl.jena.query.ResultSet;
import com.hp.hpl.jena.query.Syntax;
+import com.hp.hpl.jena.rdf.model.Model;
+import com.hp.hpl.jena.rdf.model.ModelFactory;
import com.hp.hpl.jena.rdf.model.RDFNode;
+import com.hp.hpl.jena.rdf.model.Statement;
import com.hp.hpl.jena.sparql.engine.http.QueryEngineHTTP;
import com.hp.hpl.jena.vocabulary.OWL;
import com.jamonapi.Monitor;
@@ -59,6 +86,22 @@
//KB2
private KnowledgeBase kb2;
+ private Map<Description, List<? extends EvaluatedDescription>> mappingKB1KB2;
+ private Map<Description, List<? extends EvaluatedDescription>> mappingKB2KB1;
+
+ private boolean posNegLearning = true;
+
+ /**
+ * The maximum number of positive examples, used for the SPARQL extraction and learning algorithm
+ */
+ private int maxNrOfPositiveExamples = 20;
+ /**
+ * The maximum number of negative examples, used for the SPARQL extraction and learning algorithm
+ */
+ private int maxNrOfNegativeExamples = 20;
+
+ private NamedClass currentClass;
+
public OntologyMatching(KnowledgeBase kb1, KnowledgeBase kb2) {
this.kb1 = kb1;
this.kb2 = kb2;
@@ -71,12 +114,28 @@
}
public void start(){
- Map<Description, List<? extends EvaluatedDescription>> mapping1 = computeMapping(kb1, kb2);
- printMappingPretty(mapping1);
- Map<Description, List<? extends EvaluatedDescription>> mapping2 = computeMapping(kb2, kb1);
- printMappingPretty(mapping2);
+ mappingKB1KB2 = computeAlignment(kb1, kb2);
+ printMappingPretty(mappingKB1KB2);
+ mappingKB2KB1 = computeAlignment(kb2, kb1);
+ printMappingPretty(mappingKB2KB1);
}
+ public Map<Description, List<? extends EvaluatedDescription>> getMappingKB1KB2() {
+ return mappingKB1KB2;
+ }
+
+ public Map<Description, List<? extends EvaluatedDescription>> getMappingKB2KB1() {
+ return mappingKB2KB1;
+ }
+
+ public void setMaxNrOfPositiveExamples(int maxNrOfPositiveExamples) {
+ this.maxNrOfPositiveExamples = maxNrOfPositiveExamples;
+ }
+
+ public void setMaxNrOfNegativeExamples(int maxNrOfNegativeExamples) {
+ this.maxNrOfNegativeExamples = maxNrOfNegativeExamples;
+ }
+
private void printMapping(Map<Description, List<? extends EvaluatedDescription>> mapping){
logger.info("Source Class -> Target Class Expression");
for (Entry<Description, List<? extends org.dllearner.core.EvaluatedDescription>> entry : mapping.entrySet()) {
@@ -106,78 +165,139 @@
}
List<? extends org.dllearner.core.EvaluatedDescription> value = entry.getValue();
logger.info(OWLAPIDescriptionConvertVisitor.getOWLClassExpression(key));
- for (EvaluatedDescription evaluatedDescription : value) {
- logger.info(indention + "\t->\t" +
- OWLAPIDescriptionConvertVisitor.getOWLClassExpression(evaluatedDescription.getDescription()) +
- "(" + dfPercent.format(evaluatedDescription.getAccuracy()) + ")");
+ if(value == null){
+ logger.info(indention + "\t->\t ERROR");
+ } else {
+ for (EvaluatedDescription evaluatedDescription : value) {
+ logger.info(indention + "\t->\t" +
+ OWLAPIDescriptionConvertVisitor.getOWLClassExpression(evaluatedDescription.getDescription()) +
+ "(" + dfPercent.format(evaluatedDescription.getAccuracy()) + ")");
+ }
}
+
}
}
- private Map<Description, List<? extends EvaluatedDescription>> computeMapping(KnowledgeBase source, KnowledgeBase target) {
+ private Map<Description, List<? extends EvaluatedDescription>> computeAlignment(KnowledgeBase source, KnowledgeBase target) {
Map<Description, List<? extends EvaluatedDescription>> mapping = new HashMap<Description, List<? extends EvaluatedDescription>>();
// get all classes in SOURCE
Set<NamedClass> sourceClasses = getClasses(source);
// for each class of KB1
for (NamedClass nc : sourceClasses) {
- logger.info(nc);
- // get all via owl:sameAs related individuals
- SortedSet<Individual> individuals = getRelatedIndividualsNamespaceAware(source, nc, target.getNamespace());
- logger.info(individuals.size());
- //learn concept in KB2 based on the examples
- if(individuals.size() >= 3){
- individuals = SetManipulation.fuzzyShrinkInd(individuals, 10);
- List<? extends EvaluatedDescription> learnedClassExpressions = learnClassExpression(target, individuals);
- mapping.put(nc, learnedClassExpressions);
+ try {
+ logger.info(nc);
+ List<? extends EvaluatedDescription> learnedClassExpressions = computeMapping(nc, source, target);
+ if(learnedClassExpressions != null){
+ mapping.put(nc, learnedClassExpressions);
+ //break;
+ }
+ } catch (Exception e) {
+ logger.error("Failed for " + nc.getName(), e);
}
}
return mapping;
}
- private List<? extends EvaluatedDescription> learnClassExpression(KnowledgeBase kb, SortedSet<Individual> posExamples){
- return learnClassExpression(kb, posExamples, false);
+ public List<? extends EvaluatedDescription> computeMapping(NamedClass sourceClass, KnowledgeBase source, KnowledgeBase target){
+ currentClass = sourceClass;
+ List<? extends EvaluatedDescription> learnedClassExpressions = null;
+ // get all via owl:sameAs related individuals
+ SortedSet<Individual> relatedIndividuals = getRelatedIndividualsNamespaceAware(source, sourceClass, target.getNamespace());
+ logger.info("#Resources in target KB: " + relatedIndividuals.size());
+ //learn concept in KB2 based on the examples
+ if(relatedIndividuals.size() >= 3){
+ learnedClassExpressions = learnClassExpressions(target, relatedIndividuals, posNegLearning);
+ }
+ return learnedClassExpressions;
}
- private List<? extends EvaluatedDescription> learnClassExpression(KnowledgeBase kb, SortedSet<Individual> positiveExamples, boolean posNeg){
+ private List<? extends EvaluatedDescription> learnClassExpressions(KnowledgeBase kb, SortedSet<Individual> posExamples){
+ return learnClassExpressions(kb, posExamples, false);
+ }
+
+ private List<? extends EvaluatedDescription> learnClassExpressions(KnowledgeBase kb, SortedSet<Individual> positiveExamples, boolean posNeg){
+ Model fullFragment = null;
try {
+
+ //get a sample of the positive examples
+ SortedSet<Individual> positiveExamplesSample = SetManipulation.stableShrinkInd(positiveExamples, maxNrOfPositiveExamples);
+
+ //starting from the positive examples, we first extract the fragment for them
+ logger.info("Extracting fragment for positive examples...");
+ Model positiveFragment = getFragment(positiveExamplesSample, kb);
+ logger.info("...done.");
+
+ //based on the fragment we try to find some good negative examples
SortedSet<Individual> negativeExamples = new TreeSet<Individual>();
if(posNeg){
- //find negative examples
mon.start();
+ //find the classes the positive examples are asserted to
+ Set<NamedClass> classes = new HashSet<NamedClass>();
+ ParameterizedSparqlString template = new ParameterizedSparqlString("SELECT ?type WHERE {?s a ?type.}");
+ for(Individual pos : positiveExamples){
+ template.clearParams();
+ template.setIri("s", pos.getName());
+ ResultSet rs = QueryExecutionFactory.create(template.asQuery(), positiveFragment).execSelect();
+ QuerySolution qs;
+ while(rs.hasNext()){
+ qs = rs.next();
+ if(qs.get("type").isURIResource()){
+ classes.add(new NamedClass(qs.getResource("type").getURI()));
+ }
+ }
+ }
+ System.out.println(classes);
+
+ //get the negative examples
+ for(NamedClass nc : classes){
+ Set<String> parallelClasses = kb.getSparqlHelper().getParallelClasses(nc.getName(), 5);
+ for(String parallelClass : parallelClasses){
+ negativeExamples.addAll(kb.getReasoner().getIndividuals(new NamedClass(parallelClass), 5));
+ negativeExamples.removeAll(positiveExamples);
+ }
+ }
+
AutomaticNegativeExampleFinderSPARQL2 finder = new AutomaticNegativeExampleFinderSPARQL2(kb.getEndpoint());
//TODO find negative examples
mon.stop();
logger.info("Found " + negativeExamples.size() + " negative examples in " + mon.getLastValue() + "ms.");
}
- SortedSetTuple<Individual> examples = new SortedSetTuple<Individual>(positiveExamples, negativeExamples);
+ //get a sample of the positive examples
+ SortedSet<Individual> negativeExamplesSample = SetManipulation.stableShrinkInd(negativeExamples, maxNrOfNegativeExamples);
- SparqlKnowledgeSource ks = new SparqlKnowledgeSource();
- ks.setInstances(Datastructures.individualSetToStringSet(examples.getCompleteSet()));
- ks.setUrl(kb.getEndpoint().getURL());
- ks.setDefaultGraphURIs(new TreeSet<String>(kb.getEndpoint().getDefaultGraphURIs()));
- ks.setUseLits(false);
- ks.setUseCacheDatabase(true);
- ks.setCacheDir("cache");
- ks.setRecursionDepth(2);
- ks.setCloseAfterRecursion(true);
- ks.setDissolveBlankNodes(false);
- ks.setSaveExtractedFragment(false);
- ks.init();
+ logger.info("#Positive examples: " + positiveExamplesSample.size());
+ logger.info("#Negative examples: " + negativeExamplesSample.size());
+ //create fragment for negative examples
+ logger.info("Extracting fragment for negative examples...");
+ Model negativeFragment = getFragment(negativeExamplesSample, kb);
+ logger.info("...done.");
+
+ //create fragment consisting of both
+ fullFragment = ModelFactory.createDefaultModel();
+ fullFragment.add(positiveFragment);
+ fullFragment.add(negativeFragment);
+
+ //here is the most difficult task, i.e. find a 'good' fragment of the KB on which we can learn
+ KnowledgeSource ks = convert(fullFragment);
+
+ //initialize the reasoner
AbstractReasonerComponent rc = new FastInstanceChecker(ks);
rc.init();
+ //initialize the learning problem
AbstractLearningProblem lp;
if(posNeg){
- lp = new PosNegLPStandard(rc, positiveExamples, negativeExamples);
+ lp = new PosNegLPStandard(rc, positiveExamplesSample, negativeExamplesSample);
} else {
- lp = new PosOnlyLP(rc, positiveExamples);
-
+ lp = new PosOnlyLP(rc, positiveExamplesSample);
}
lp.init();
+ //apply the learning algorithm
+ logger.info("Running learning algorithm...");
CELOE la = new CELOE(lp, rc);
la.setMaxExecutionTimeInSeconds(10);
la.setNoisePercentage(25);
@@ -189,10 +309,104 @@
return la.getCurrentlyBestEvaluatedDescriptions(10);
} catch (ComponentInitException e) {
e.printStackTrace();
+ try {
+ new File("errors").mkdir();
+ fullFragment.write(new FileOutputStream("errors/" + prettyPrint(currentClass) + ".ttl"), "TURTLE", null);
+ } catch (FileNotFoundException e1) {
+ e1.printStackTrace();
+ }
}
return null;
}
+ private KnowledgeSource convert(Model model){
+ try {
+ ByteArrayOutputStream baos = new ByteArrayOutputStream();
+ model.write(baos, "TURTLE", null);
+ OWLOntologyManager man = OWLManager.createOWLOntologyManager();
+ OWLOntology ontology = man.loadOntologyFromOntologyDocument(new ByteArrayInputStream(baos.toByteArray()));
+ return new OWLAPIOntology(ontology);
+ } catch (OWLOntologyCreationException e) {
+ e.printStackTrace();
+ try {
+ model.write(new FileOutputStream(prettyPrint(currentClass) + "_conversion_error.ttl"), "TURTLE", null);
+ } catch (FileNotFoundException e1) {
+ e.printStackTrace();
+ }
+ }
+ return null;
+ }
+
+ /**
+ * Computes a fragment containing hopefully useful information about the resources.
+ * @param ind
+ */
+ private KnowledgeSource getFragmentLegacy(SortedSet<Individual> positiveExamples, SortedSet<Individual> negativeExamples, SparqlEndpoint endpoint){
+ SortedSetTuple<Individual> examples = new SortedSetTuple<Individual>(positiveExamples, negativeExamples);
+ SparqlKnowledgeSource ks = new SparqlKnowledgeSource();
+ ks.setInstances(Datastructures.individualSetToStringSet(examples.getCompleteSet()));
+ ks.setUrl(endpoint.getURL());
+ ks.setDefaultGraphURIs(new TreeSet<String>(endpoint.getDefaultGraphURIs()));
+ ks.setUseLits(false);
+ ks.setUseCacheDatabase(true);
+ ks.setCacheDir("cache");
+ ks.setRecursionDepth(2);
+ ks.setCloseAfterRecursion(true);
+ ks.setDissolveBlankNodes(false);
+ ks.setSaveExtractedFragment(false);
+ ks.init();
+ return ks;
+ }
+
+ /**
+ * Computes a fragment containing hopefully useful information about the resources.
+ * @param ind
+ */
+ private Model getFragment(SortedSet<Individual> positiveExamples, SortedSet<Individual> negativeExamples, KnowledgeBase kb){
+ OntModel fullFragment = ModelFactory.createOntologyModel();
+ int i = 1;
+ int size = Sets.union(positiveExamples, negativeExamples).size();
+ for(Individual ind : Sets.union(positiveExamples, negativeExamples)){
+ logger.info(i++ + "/" + size);
+ fullFragment.add(getFragment(ind, kb));
+ }
+ //filter out triples with String literals, as there often occur are some syntax errors and they are not relevant for learning
+ List<Statement> statementsToRemove = new ArrayList<Statement>();
+ for(Iterator<Statement> iter = fullFragment.listStatements().toList().iterator(); iter.hasNext();){
+ Statement st = iter.next();
+ RDFNode object = st.getObject();
+ if(object.isLiteral()){
+ statementsToRemove.add(st);
+// Literal lit = object.asLiteral();
+// if(lit.getDatatype() == null || lit.getDatatype().equals(XSD.STRING)){
+// iter.remove();
+// }
+ }
+ }
+ fullFragment.remove(statementsToRemove);
+ return fullFragment;
+ }
+
+ /**
+ * Computes a fragment containing hopefully useful information about the resources.
+ * @param ind
+ */
+ private Model getFragment(SortedSet<Individual> examples, KnowledgeBase kb){
+ return getFragment(examples, new TreeSet<Individual>(), kb);
+ }
+
+ /**
+ * Computes a fragment containing hopefully useful information about the resource.
+ * @param ind
+ */
+ private Model getFragment(Individual ind, KnowledgeBase kb){
+ logger.debug("Loading fragment for " + ind.getName());
+ ConciseBoundedDescriptionGenerator cbdGen = new ConciseBoundedDescriptionGeneratorImpl(kb.getEndpoint(), kb.getCache());
+ Model cbd = cbdGen.getConciseBoundedDescription(ind.getName(), 2);
+ logger.debug("Got " + cbd.size() + " triples.");
+ return cbd;
+ }
+
private Set<NamedClass> getClasses(KnowledgeBase kb){
Set<NamedClass> classes = kb.getSparqlHelper().getAllClasses();
//fallback: check for ?s a ?type where ?type is not asserted to owl:Class
@@ -282,6 +496,84 @@
return rs;
}
+ private String prettyPrint(Description desc){
+ return OWLAPIDescriptionConvertVisitor.getOWLClassExpression(desc).toString();
+ }
+
+ public static String toHTML(Map<Description, List<? extends EvaluatedDescription>> mapping){
+ StringBuilder sb = new StringBuilder();
+ DecimalFormat dfPercent = new DecimalFormat("0.00%");
+ sb.append("<html>\n");
+ sb.append("<table>\n");
+ sb.append("<thead><tr><th>Source Class</th><th>Target Class Expressions</th></tr></thead>\n");
+ sb.append("<tbody>\n");
+
+
+ for (Entry<Description, List<? extends org.dllearner.core.EvaluatedDescription>> entry : mapping.entrySet()) {
+ Description key = entry.getKey();
+ List<? extends org.dllearner.core.EvaluatedDescription> value = entry.getValue();
+ if(value == null){
+ sb.append("<tr><th>" + OWLAPIDescriptionConvertVisitor.getOWLClassExpression(key) + "</th>\n");
+ sb.append("<tr><td>ERROR</td></tr>\n");
+ } else {
+ sb.append("<tr><th rowspan=\"" + value.size()+1 + "\">" + OWLAPIDescriptionConvertVisitor.getOWLClassExpression(key) + "</th>\n");
+ for (EvaluatedDescription evaluatedDescription : value) {
+ sb.append("<tr><td>" +
+ OWLAPIDescriptionConvertVisitor.getOWLClassExpression(evaluatedDescription.getDescription()) + "(" + dfPercent.format(evaluatedDescription.getAccuracy()) + ")"
+ + "</td></tr>\n");
+ }
+ }
+ }
+
+ sb.append("</tbody>\n");
+ sb.append("</table>\n");
+ sb.append("</html>\n");
+
+ return sb.toString();
+ }
+
+ public static String toHTMLWithLabels(Map<Description, List<? extends EvaluatedDescription>> mapping, KnowledgeBase source, KnowledgeBase target){
+ ManchesterOWLSyntaxOWLObjectRendererImpl sourceRenderer = new ManchesterOWLSyntaxOWLObjectRendererImpl();
+ sourceRenderer.setShortFormProvider(new LabelShortFormProvider(source.getEndpoint(), source.getCache()));
+ ManchesterOWLSyntaxOWLObjectRendererImpl targetRenderer = new ManchesterOWLSyntaxOWLObjectRendererImpl();
+ targetRenderer.setShortFormProvider(new LabelShortFormProvider(target.getEndpoint(), target.getCache()));
+
+ StringBuilder sb = new StringBuilder();
+ DecimalFormat dfPercent = new DecimalFormat("0.00%");
+ sb.append("<html>\n");
+ sb.append("<table>\n");
+ sb.append("<thead><tr><th>Source Class</th><th>Target Class Expressions</th><th>Accuracy</th></tr></thead>\n");
+ sb.append("<tbody>\n");
+
+
+ for (Entry<Description, List<? extends org.dllearner.core.EvaluatedDescription>> entry : mapping.entrySet()) {
+ Description key = entry.getKey();
+ String renderedKey = sourceRenderer.render(OWLAPIDescriptionConvertVisitor.getOWLClassExpression(key));
+ List<? extends org.dllearner.core.EvaluatedDescription> value = entry.getValue();
+ if(value == null){
+ sb.append("<tr><th>" + renderedKey + "</th>\n");
+ sb.append("<tr><td>ERROR</td><td></td></tr>\n");
+ } else {
+ sb.append("<tr><th rowspan=\"" + (value.size()+1) + "\">" + renderedKey + "</th>\n");
+ for (EvaluatedDescription evaluatedDescription : value) {
+ sb.append("<tr>");
+ String renderedDesc = targetRenderer.render(OWLAPIDescriptionConvertVisitor.getOWLClassExpression(evaluatedDescription.getDescription()));
+ sb.append("<td>" + renderedDesc + "</td>");
+ sb.append("<td>" + dfPercent.format(evaluatedDescription.getAccuracy()) + "</td>");
+ sb.append("</tr>\n");
+ }
+ }
+ }
+
+ sb.append("</tbody>\n");
+ sb.append("</table>\n");
+ sb.append("</html>\n");
+
+ return sb.toString();
+ }
+
+
+
public static class KnowledgeBase{
private SparqlEndpoint endpoint;
private SPARQLReasoner reasoner;
@@ -294,7 +586,7 @@
this.namespace = namespace;
this.cache = cache;
- this.reasoner = new SPARQLReasoner(new SparqlEndpointKS(endpoint));
+ this.reasoner = new SPARQLReasoner(new SparqlEndpointKS(endpoint), cache);
this.sparqlHelper = new SPARQLTasks(endpoint);
}
Modified: trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatchingTest.java
===================================================================
--- trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatchingTest.java 2013-02-06 17:59:21 UTC (rev 3894)
+++ trunk/scripts/src/main/java/org/dllearner/scripts/OntologyMatchingTest.java 2013-02-06 18:01:42 UTC (rev 3895)
@@ -1,11 +1,20 @@
package org.dllearner.scripts;
+import java.io.BufferedWriter;
+import java.io.FileWriter;
+import java.io.IOException;
import java.net.URL;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
import org.apache.log4j.ConsoleAppender;
import org.apache.log4j.Level;
import org.apache.log4j.Logger;
import org.apache.log4j.PatternLayout;
+import org.dllearner.core.EvaluatedDescription;
+import org.dllearner.core.owl.Description;
+import org.dllearner.core.owl.NamedClass;
import org.dllearner.kb.extraction.ExtractionAlgorithm;
import org.dllearner.kb.sparql.ExtractionDBCache;
import org.dllearner.kb.sparql.SparqlEndpoint;
@@ -34,12 +43,13 @@
Logger.getLogger(org.dllearner.kb.extraction.Manager.class).setLevel(Level.WARN);
Logger.getRootLogger().removeAllAppenders();
Logger.getRootLogger().addAppender(new ConsoleAppender(new PatternLayout("%m%n")));
+// Logger.getRootLogger().setLevel(Level.DEBUG);
//DBpedia
SparqlEndpoint endpoint = SparqlEndpoint.getEndpointDBpedia();
ExtractionDBCache cache = new ExtractionDBCache("cache");
String namespace = "http://dbpedia.org/resource/";
- dbpedia = new KnowledgeBase(endpoint, cache, namespace);
+ dbpedia = new KnowledgeBase(endpoint, null, namespace);
//World Factbook
//TODO problem with World Factbook is that old FU Berlin server is useless because of bugs and current version
@@ -55,31 +65,83 @@
endpoint = new SparqlEndpoint(new URL("http://localhost:8890/sparql"));
cache = new ExtractionDBCache("cache");
namespace = "http://sw.cyc.com";
- openCyc = new KnowledgeBase(endpoint, cache, namespace);
+ openCyc = new KnowledgeBase(endpoint, null, namespace);
//LinkedGeoData
endpoint = new SparqlEndpoint(new URL("http://linkedgeodata.org/sparql"));
cache = new ExtractionDBCache("cache");
namespace = "http://linkedgeodata.org/triplify/";
- linkedGeoData = new KnowledgeBase(endpoint, cache, namespace);
+ linkedGeoData = new KnowledgeBase(endpoint, null, namespace);
}
@Test
public void testDBpediaWorldFactbook() {
OntologyMatching matcher = new OntologyMatching(dbpedia, worldFactBook);
matcher.start();
+ save(matcher.getMappingKB1KB2(), "dbpedia_wfb.html");
+ save(matcher.getMappingKB2KB1(), "wfb_dbpedia.html");
}
@Test
public void testDBpediaOpenCyc() {
OntologyMatching matcher = new OntologyMatching(dbpedia, openCyc);
matcher.start();
+ save("dbpedia_opencyc.html", matcher.getMappingKB1KB2(), dbpedia, openCyc);
+ save("opencyc_dbpedia.html", matcher.getMappingKB2KB1(), openCyc, dbpedia);
}
@Test
public void testDBpediaLinkedGeoData() {
OntologyMatching matcher = new OntologyMatching(dbpedia, linkedGeoData);
matcher.start();
+ save("dbpedia_lgd.html", matcher.getMappingKB1KB2(), dbpedia, linkedGeoData );
+ save("lgd_dbpedia.html", matcher.getMappingKB2KB1(), linkedGeoData, dbpedia);
}
+
+ @Test
+ public void testSingleClass() {
+ OntologyMatching matcher = new OntologyMatching(openCyc, dbpedia);
+ NamedClass nc = new NamedClass("http://sw.opencyc.org/concept/Mx4r4fYeXvbPQdiKtoNafhmOew");
+ List<? extends EvaluatedDescription> mapping = matcher.computeMapping(nc, openCyc, dbpedia);
+ Map<Description, List<? extends EvaluatedDescription>> alignment = new HashMap<Description, List<? extends EvaluatedDescription>>();
+ alignment.put(nc, mapping);
+ System.out.println(OntologyMatching.toHTMLWithLabels(alignment, openCyc, dbpedia));
+ }
+
+ private void save(String filename, Map<Description, List<? extends EvaluatedDescription>> mapping, KnowledgeBase source, KnowledgeBase target){
+ BufferedWriter out = null;
+ try {
+ out = new BufferedWriter(new FileWriter(filename));
+ out.write(OntologyMatching.toHTMLWithLabels(mapping, source, target));
+ } catch (IOException e) {
+ e.printStackTrace();
+ } finally {
+ if (out != null) {
+ try {
+ out.close();
+ } catch (IOException e) {
+ e.printStackTrace();
+ }
+ }
+ }
+ }
+
+ private void save(Map<Description, List<? extends EvaluatedDescription>> mapping, String filename){
+ BufferedWriter out = null;
+ try {
+ out = new BufferedWriter(new FileWriter(filename));
+ out.write(OntologyMatching.toHTML(mapping));
+ } catch (IOException e) {
+ e.printStackTrace();
+ } finally {
+ if (out != null) {
+ try {
+ out.close();
+ } catch (IOException e) {
+ e.printStackTrace();
+ }
+ }
+ }
+ }
}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X