Menu
▾
▴
[DL-Learner SVN] SF.net SVN: dl-learner:[3322] trunk/components-ext/src/main/java/org/ dllearner/algorithm/tbsl
|
From: <lor...@us...> - 2011-10-26 09:04:10
|
Revision: 3322
http://dl-learner.svn.sourceforge.net/dl-learner/?rev=3322&view=rev
Author: lorenz_b
Date: 2011-10-26 09:03:57 +0000 (Wed, 26 Oct 2011)
Log Message:
-----------
Some changes to get fuzzy search working.
Added some methods to Tagger for POS Tagger evlauation.
Modified Paths:
--------------
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/learning/SPARQLTemplateBasedLearner.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/ltag/parser/Preprocessor.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/ApachePartOfSpeechTagger.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/DBpediaSpotlightNER.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/LingPipePartOfSpeechTagger.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/PartOfSpeechTagger.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/StanfordPartOfSpeechTagger.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/TreeTagger.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/WordNet.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/search/SolrSearch.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/search/ThresholdSlidingSolrSearch.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/Query.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/SPARQL_Term.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/Template.java
trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/templator/Templator.java
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/learning/SPARQLTemplateBasedLearner.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/learning/SPARQLTemplateBasedLearner.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/learning/SPARQLTemplateBasedLearner.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -27,6 +27,7 @@
import org.dllearner.algorithm.tbsl.sparql.Query;
import org.dllearner.algorithm.tbsl.sparql.RatedQuery;
import org.dllearner.algorithm.tbsl.sparql.SPARQL_Prefix;
+import org.dllearner.algorithm.tbsl.sparql.SPARQL_QueryType;
import org.dllearner.algorithm.tbsl.sparql.Slot;
import org.dllearner.algorithm.tbsl.sparql.SlotType;
import org.dllearner.algorithm.tbsl.sparql.Template;
@@ -48,6 +49,7 @@
import com.hp.hpl.jena.query.ResultSet;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.rdf.model.ModelFactory;
+import com.hp.hpl.jena.sparql.engine.http.QueryEngineHTTP;
import com.jamonapi.Monitor;
import com.jamonapi.MonitorFactory;
@@ -67,7 +69,7 @@
private Ranking ranking;
private boolean useRemoteEndpointValidation;
private boolean stopIfQueryResultNotEmpty;
- private int maxTestedQueriesPerTemplate;
+ private int maxTestedQueriesPerTemplate = 50;
private int maxQueryExecutionTimeInSeconds;
private SparqlEndpoint endpoint = SparqlEndpoint.getEndpointDBpediaLiveAKSW();
@@ -88,7 +90,7 @@
private Map<String, List<String>> classesURICache;
private Map<String, List<String>> propertiesURICache;
- private Map<String, List<String>> learnedSPARQLQueries;
+ private Map<String, Object> learnedSPARQLQueries;
private Set<Template> templates;
private Collection<Query> sparqlQueryCandidates;
private Map<Template, Collection<? extends Query>> template2Queries;
@@ -120,6 +122,13 @@
templateGenerator = new Templator();
}
+ /*
+ * Only for Evaluation useful.
+ */
+ public void setUseIdealTagger(boolean value){
+ templateGenerator.setUNTAGGED_INPUT(!value);
+ }
+
private void init(Options options){
String resourcesIndexUrl = options.fetch("solr.resources.url");
String resourcesIndexSearchField = options.fetch("solr.resources.searchfield");
@@ -127,8 +136,14 @@
String classesIndexUrl = options.fetch("solr.classes.url");
String classesIndexSearchField = options.fetch("solr.classes.searchfield");
- class_index = new ThresholdSlidingSolrSearch(classesIndexUrl, classesIndexSearchField, 1.0, 0.1);
+ SolrSearch dbpediaClassIndex = new SolrSearch(classesIndexUrl, classesIndexSearchField);
+ String yagoClassesIndexUrl = options.fetch("solr.yago.classes.url");
+ String yagoClassesIndexSearchField = options.fetch("solr.yago.classes.searchfield");
+ SolrSearch yagoClassIndex = new SolrSearch(yagoClassesIndexUrl, yagoClassesIndexSearchField);
+
+ class_index = new ThresholdSlidingSolrSearch(dbpediaClassIndex);// new HierarchicalSolrSearch(dbpediaClassIndex, yagoClassIndex);
+
String propertiesIndexUrl = options.fetch("solr.properties.url");
String propertiesIndexSearchField = options.fetch("solr.properties.searchfield");
SolrSearch labelBasedPropertyIndex = new SolrSearch(propertiesIndexUrl, propertiesIndexSearchField);
@@ -137,8 +152,12 @@
String boaPatternIndexSearchField = options.fetch("solr.boa.properties.searchfield");
SolrSearch patternBasedPropertyIndex = new SolrSearch(boaPatternIndexUrl, boaPatternIndexSearchField);
- property_index = new HierarchicalSolrSearch(patternBasedPropertyIndex, labelBasedPropertyIndex);
+ //first BOA pattern then label based
+// property_index = new HierarchicalSolrSearch(patternBasedPropertyIndex, labelBasedPropertyIndex);
+ //first label based then BOA pattern
+ property_index = new HierarchicalSolrSearch(labelBasedPropertyIndex, patternBasedPropertyIndex);
+
int maxIndexResults = Integer.parseInt(options.fetch("solr.query.limit"), 10);
maxQueryExecutionTimeInSeconds = Integer.parseInt(options.get("sparql.query.maxExecutionTimeInSeconds", "20"));
@@ -191,7 +210,7 @@
}
private void reset(){
- learnedSPARQLQueries = new HashMap<String, List<String>>();
+ learnedSPARQLQueries = new HashMap<String, Object>();
resourcesURICache = new HashMap<String, List<String>>();
classesURICache = new HashMap<String, List<String>>();
propertiesURICache = new HashMap<String, List<String>>();
@@ -417,13 +436,13 @@
SPARQL_Prefix prefix = null;
uriCandidates = getCandidateURIsSortedBySimilarity(slot);
for(String uri : uriCandidates){
- for(Entry<String, String> uri2prefix : prefixMap.entrySet()){
- if(uri.startsWith(uri2prefix.getKey())){
- prefix = new SPARQL_Prefix(uri2prefix.getValue(), uri2prefix.getKey());
- uri = uri.replace(uri2prefix.getKey(), uri2prefix.getValue() + ":");
- break;
- }
- }
+// for(Entry<String, String> uri2prefix : prefixMap.entrySet()){
+// if(uri.startsWith(uri2prefix.getKey())){
+// prefix = new SPARQL_Prefix(uri2prefix.getValue(), uri2prefix.getKey());
+// uri = uri.replace(uri2prefix.getKey(), uri2prefix.getValue() + ":");
+// break;
+// }
+// }
for(Query query : queries){
if(slot.getSlotType() == SlotType.SYMPROPERTY){
Query reversedQuery = new Query(query);
@@ -635,30 +654,57 @@
private void validateAgainstRemoteEndpoint(Collection<? extends Query> queries){
List<String> queryStrings = new ArrayList<String>();
+ SPARQL_QueryType queryType = SPARQL_QueryType.SELECT;
for(Query query : queries){
+ if(query.getQt() == SPARQL_QueryType.ASK){
+ queryType = SPARQL_QueryType.ASK;
+ } else if(query.getQt() == SPARQL_QueryType.SELECT){
+ queryType = SPARQL_QueryType.SELECT;
+ }
queryStrings.add(query.toString());
}
- validateAgainstRemoteEndpoint(queryStrings);
+ validateAgainstRemoteEndpoint(queryStrings, queryType);
}
- private void validateAgainstRemoteEndpoint(List<String> queries){
+ private void validateAgainstRemoteEndpoint(List<String> queries, SPARQL_QueryType queryType){
logger.info("Testing candidate SPARQL queries on remote endpoint...");
mon.start();
- for(String query : queries){
- logger.info("Testing query:\n" + query);
- List<String> results = getResultFromRemoteEndpoint(query);
- if(!results.isEmpty()){
- learnedSPARQLQueries.put(query, results);
- if(stopIfQueryResultNotEmpty){
+ if(queryType == SPARQL_QueryType.SELECT){
+ for(String query : queries){
+ logger.info("Testing query:\n" + query);
+ List<String> results = getResultFromRemoteEndpoint(query);
+ if(!results.isEmpty()){
+ learnedSPARQLQueries.put(query, results);
+ if(stopIfQueryResultNotEmpty){
+ return;
+ }
+ }
+ logger.info("Result: " + results);
+ }
+ } else if(queryType == SPARQL_QueryType.ASK){
+ for(String query : queries){
+ logger.info("Testing query:\n" + query);
+ boolean result = executeAskQuery(query);
+ learnedSPARQLQueries.put(query, result);
+ if(stopIfQueryResultNotEmpty && result){
return;
}
+ logger.info("Result: " + result);
}
- logger.info("Result: " + results);
}
mon.stop();
logger.info("Done in " + mon.getLastValue() + "ms.");
}
+ private boolean executeAskQuery(String query){
+ QueryEngineHTTP qe = new QueryEngineHTTP(endpoint.getURL().toString(), query);
+ for(String uri : endpoint.getDefaultGraphURIs()){
+ qe.addDefaultGraph(uri);
+ }
+ boolean ret = qe.execAsk();
+ return ret;
+ }
+
private void validateAgainstLocalModel(Collection<? extends Query> queries){
List<String> queryStrings = new ArrayList<String>();
for(Query query : queries){
@@ -686,7 +732,11 @@
private List<String> getResultFromRemoteEndpoint(String query){
List<String> resources = new ArrayList<String>();
try {
- ResultSet rs = SparqlQuery.convertJSONtoResultSet(cache.executeSelectQuery(endpoint, query + " LIMIT 10"));
+ String queryString = query;
+ if(!query.contains("LIMIT") && !query.contains("ASK")){
+ queryString = query + " LIMIT 10";
+ }
+ ResultSet rs = SparqlQuery.convertJSONtoResultSet(cache.executeSelectQuery(endpoint, queryString));
QuerySolution qs;
String projectionVar;
while(rs.hasNext()){
@@ -724,8 +774,7 @@
// Logger.getLogger(DefaultHttpParams.class).setLevel(Level.OFF);
// Logger.getLogger(HttpClient.class).setLevel(Level.OFF);
// Logger.getLogger(HttpMethodBase.class).setLevel(Level.OFF);
-// String question = "Give me all books written by authors influenced by Ernest Hemingway.";
- String question = "Give me all European Capitals!";
+ String question = "Who wrote the book The pillars of the Earth?";
// String question = "Give me all books written by authors influenced by Ernest Hemingway.";
SPARQLTemplateBasedLearner learner = new SPARQLTemplateBasedLearner();
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/ltag/parser/Preprocessor.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/ltag/parser/Preprocessor.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/ltag/parser/Preprocessor.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -7,6 +7,7 @@
import java.util.regex.Pattern;
import org.apache.log4j.Logger;
+import org.dllearner.algorithm.tbsl.nlp.DBpediaSpotlightNER;
import org.dllearner.algorithm.tbsl.nlp.LingPipeNER;
import org.dllearner.algorithm.tbsl.nlp.NER;
import org.dllearner.algorithm.tbsl.templator.Templator;
@@ -24,6 +25,7 @@
USE_NER = n;
if (USE_NER) {
ner = new LingPipeNER(true); //not case sensitive best solution?
+// ner = new DBpediaSpotlightNER();
}
}
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/ApachePartOfSpeechTagger.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/ApachePartOfSpeechTagger.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/ApachePartOfSpeechTagger.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -3,12 +3,15 @@
import java.io.IOException;
import java.io.InputStream;
import java.util.ArrayList;
+import java.util.Arrays;
import java.util.List;
import opennlp.tools.postag.POSModel;
import opennlp.tools.postag.POSTaggerME;
import opennlp.tools.util.Sequence;
+import com.aliasi.tag.Tagging;
+
public class ApachePartOfSpeechTagger implements PartOfSpeechTagger{
private POSTaggerME tagger;
@@ -41,6 +44,11 @@
tokenizer = new ApacheTokenizer();
}
+
+ @Override
+ public String getName() {
+ return "Apache Open NLP POS Tagger";
+ }
@Override
public String tag(String sentence) {
@@ -49,6 +57,13 @@
return convert2TaggedSentence(tokens, tags);
}
+
+ public List<String> getTags(String sentence){
+ String[] tokens = tokenizer.tokenize(sentence);
+ String[] tags = tagger.tag(tokens);
+
+ return Arrays.asList(tags);
+ }
@Override
public List<String> tagTopK(String sentence) {
@@ -61,6 +76,14 @@
return taggedSentences;
}
+ @Override
+ public Tagging<String> getTagging(String sentence){
+ String[] tokens = tokenizer.tokenize(sentence);
+ String[] tags = tagger.tag(tokens);
+
+ return new Tagging<String>(Arrays.asList(tokens), Arrays.asList(tags));
+ }
+
private String convert2TaggedSentence(String[] words, String[] tags){
StringBuilder sb = new StringBuilder();
for(int i = 0; i < words.length; i++){
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/DBpediaSpotlightNER.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/DBpediaSpotlightNER.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/DBpediaSpotlightNER.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -48,15 +48,17 @@
}
rd.close();
JSONObject json = new JSONObject(sb.toString());
- JSONArray array = json.getJSONArray("Resources");
- JSONObject entityObject;
- for(int i = 0; i < array.length(); i++){
- entityObject = array.getJSONObject(i);
- System.out.println("Entity: " + entityObject.getString("@surfaceForm"));
- System.out.println("DBpedia URI: " + entityObject.getString("@URI"));
- System.out.println("Types: " + entityObject.getString("@types"));
- namedEntities.add(entityObject.getString("@surfaceForm"));
-
+ if(!json.isNull("Resources")){
+ JSONArray array = json.getJSONArray("Resources");
+ JSONObject entityObject;
+ for(int i = 0; i < array.length(); i++){
+ entityObject = array.getJSONObject(i);
+ System.out.println("Entity: " + entityObject.getString("@surfaceForm"));
+ System.out.println("DBpedia URI: " + entityObject.getString("@URI"));
+ System.out.println("Types: " + entityObject.getString("@types"));
+ namedEntities.add(entityObject.getString("@surfaceForm"));
+
+ }
}
} catch (MalformedURLException e) {
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/LingPipePartOfSpeechTagger.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/LingPipePartOfSpeechTagger.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/LingPipePartOfSpeechTagger.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -42,6 +42,11 @@
}
@Override
+ public String getName() {
+ return "LingPipe POS Tagger";
+ }
+
+ @Override
public String tag(String sentence) {
com.aliasi.tokenizer.Tokenizer tokenizer = IndoEuropeanTokenizerFactory.INSTANCE.tokenizer(sentence.toCharArray(), 0, sentence.length());
// Tokenizer tokenizer = TOKENIZER_FACTORY.tokenizer(cs,0,cs.length);
@@ -66,5 +71,16 @@
}
return taggedSentences;
}
+
+ @Override
+ public Tagging<String> getTagging(String sentence) {
+ com.aliasi.tokenizer.Tokenizer tokenizer = IndoEuropeanTokenizerFactory.INSTANCE.tokenizer(sentence.toCharArray(), 0, sentence.length());
+// Tokenizer tokenizer = TOKENIZER_FACTORY.tokenizer(cs,0,cs.length);
+ String[] tokens = tokenizer.tokenize();
+ List<String> tokenList = Arrays.asList(tokens);
+ Tagging<String> tagging = tagger.tag(tokenList);
+
+ return tagging;
+ }
}
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/PartOfSpeechTagger.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/PartOfSpeechTagger.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/PartOfSpeechTagger.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -2,10 +2,17 @@
import java.util.List;
+import com.aliasi.tag.Tagging;
+
public interface PartOfSpeechTagger {
+ String getName();
+
String tag(String sentence);
List<String> tagTopK(String sentence);
+
+ Tagging<String> getTagging(String sentence);
+
}
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/StanfordPartOfSpeechTagger.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/StanfordPartOfSpeechTagger.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/StanfordPartOfSpeechTagger.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -4,9 +4,12 @@
import java.io.StringReader;
import java.net.URISyntaxException;
import java.util.ArrayList;
+import java.util.Arrays;
import java.util.Collections;
import java.util.List;
+import com.aliasi.tag.Tagging;
+
import edu.stanford.nlp.ling.HasWord;
import edu.stanford.nlp.ling.TaggedWord;
import edu.stanford.nlp.tagger.maxent.MaxentTagger;
@@ -29,6 +32,11 @@
e.printStackTrace();
}
}
+
+ @Override
+ public String getName() {
+ return "Stanford POS Tagger";
+ }
@Override
public String tag(String sentence) {
@@ -53,5 +61,46 @@
public List<String> tagTopK(String sentence) {
return Collections.singletonList(tag(sentence));
}
+
+ public List<String> getTags(String sentence){
+ List<String> tags = new ArrayList<String>();
+
+ ArrayList<TaggedWord> tagged = new ArrayList<TaggedWord>();
+
+ StringReader reader = new StringReader(sentence);
+ List<List<HasWord>> text = MaxentTagger.tokenizeText(reader);
+
+ if (text.size() == 1) {
+ tagged = tagger.tagSentence(text.get(0));
+ }
+
+ for(TaggedWord tW : tagged){
+ tags.add(tW.tag());
+ }
+
+ return tags;
+ }
+
+ @Override
+ public Tagging<String> getTagging(String sentence){
+ ArrayList<TaggedWord> tagged = new ArrayList<TaggedWord>();
+
+ StringReader reader = new StringReader(sentence);
+ List<List<HasWord>> text = MaxentTagger.tokenizeText(reader);
+
+ if (text.size() == 1) {
+ tagged = tagger.tagSentence(text.get(0));
+ }
+
+ List<String> tokenList = new ArrayList<String>();
+ List<String> tagList = new ArrayList<String>();
+
+ for(TaggedWord tW : tagged){
+ tokenList.add(tW.word());
+ tagList.add(tW.tag());
+ }
+
+ return new Tagging<String>(tokenList, tagList);
+ }
}
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/TreeTagger.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/TreeTagger.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/TreeTagger.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -8,6 +8,8 @@
import org.annolab.tt4j.TreeTaggerException;
import org.annolab.tt4j.TreeTaggerWrapper;
+import com.aliasi.tag.Tagging;
+
public class TreeTagger implements PartOfSpeechTagger {
TreeTaggerWrapper<String> tt;
@@ -40,6 +42,17 @@
}
return "";
}
+
+ @Override
+ public String getName() {
+ return "Tree Tagger";
+ }
+
+ @Override
+ public Tagging<String> getTagging(String sentence) {
+ // TODO Auto-generated method stub
+ return null;
+ }
@Override
public List<String> tagTopK(String sentence) {
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/WordNet.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/WordNet.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/nlp/WordNet.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -8,6 +8,9 @@
import net.didion.jwnl.JWNLException;
import net.didion.jwnl.data.IndexWord;
import net.didion.jwnl.data.POS;
+import net.didion.jwnl.data.Pointer;
+import net.didion.jwnl.data.PointerTarget;
+import net.didion.jwnl.data.PointerType;
import net.didion.jwnl.data.PointerUtils;
import net.didion.jwnl.data.Synset;
import net.didion.jwnl.data.Word;
@@ -52,6 +55,24 @@
return synonyms;
}
+ public List<String> getSisterTerms(POS pos, String s){
+ List<String> sisterTerms = new ArrayList<String>();
+
+ try {
+ IndexWord iw = dict.getIndexWord(pos, s);//dict.getMorphologicalProcessor().lookupBaseForm(pos, s)
+// IndexWord iw = dict.getMorphologicalProcessor().lookupBaseForm(pos, s);
+ if(iw != null){
+ Synset[] synsets = iw.getSenses();System.out.println(synsets[0]);
+ PointerTarget[] pointerArr = synsets[0].getTargets();
+ System.out.println(pointerArr);
+ }
+
+ } catch (JWNLException e) {
+ e.printStackTrace();
+ }
+ return sisterTerms;
+ }
+
public List<String> getAttributes(String s) {
List<String> result = new ArrayList<String>();
@@ -78,6 +99,7 @@
public static void main(String[] args) {
System.out.println(new WordNet().getBestSynonyms(POS.VERB, "learn"));
+ System.out.println(new WordNet().getSisterTerms(POS.NOUN, "actress"));
}
/**
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/search/SolrSearch.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/search/SolrSearch.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/search/SolrSearch.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -41,6 +41,14 @@
this(solrServerURL);
this.searchField = searchField;
}
+
+ public String getServerURL() {
+ return server.getBaseURL();
+ }
+
+ public String getSearchField() {
+ return searchField;
+ }
@Override
public List<String> getResources(String queryString) {
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/search/ThresholdSlidingSolrSearch.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/search/ThresholdSlidingSolrSearch.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/search/ThresholdSlidingSolrSearch.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -1,13 +1,17 @@
package org.dllearner.algorithm.tbsl.search;
+import java.text.NumberFormat;
import java.util.ArrayList;
import java.util.List;
+import java.util.Locale;
public class ThresholdSlidingSolrSearch extends SolrSearch {
private double minThreshold = 0.8;
private double step = 0.1;
+ private NumberFormat format = NumberFormat.getInstance(Locale.GERMAN);
+
public ThresholdSlidingSolrSearch(String solrServerURL) {
super(solrServerURL);
}
@@ -22,26 +26,29 @@
this.step = step;
}
+ public ThresholdSlidingSolrSearch(SolrSearch search){
+ super(search.getServerURL(), search.getSearchField());
+ }
+
+
@Override
public List<String> getResources(String queryString, int limit, int offset) {
List<String> resources = new ArrayList<String>();
+
double threshold = 1;
String queryWithThreshold = queryString;
while(resources.size() < limit && threshold >= minThreshold){
if(threshold < 1){
- queryWithThreshold = queryString + "~" + threshold;
+ queryWithThreshold = queryString + "~" + format.format(threshold);
}
resources.addAll(findResources(queryWithThreshold, limit - resources.size(), 0));
threshold -= step;
}
-
return resources;
}
-
-
}
\ No newline at end of file
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/Query.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/Query.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/Query.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -49,6 +49,7 @@
//copy constructor
public Query(Query query){
+ this.qt = query.getQt();
Set<SPARQL_Term> selTerms = new HashSet<SPARQL_Term>();
for(SPARQL_Term term : query.getSelTerms()){
SPARQL_Term newTerm = new SPARQL_Term(term.getName());
@@ -56,6 +57,7 @@
newTerm.setIsURI(newTerm.isURI);
newTerm.setAggregate(term.getAggregate());
newTerm.setOrderBy(term.getOrderBy());
+ newTerm.setAlias(term.getAlias());
selTerms.add(newTerm);
}
this.selTerms = selTerms;
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/SPARQL_Term.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/SPARQL_Term.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/SPARQL_Term.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -73,6 +73,14 @@
public void setIsURI(boolean isURI){
this.isURI = isURI;
}
+
+ public String getAlias() {
+ return alias;
+ }
+
+ public void setAlias(String alias) {
+ this.alias = alias;
+ }
@Override
public String toString() {
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/Template.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/Template.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/sparql/Template.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -39,10 +39,12 @@
}
public List<String> getLexicalAnswerType(){
- String variable = query.getAnswerTypeVariable();
- for(Slot slot : slots){
- if(slot.getAnchor().equals(variable)){
- return slot.getWords();
+ if(query.getQt() == SPARQL_QueryType.SELECT){
+ String variable = query.getAnswerTypeVariable();
+ for(Slot slot : slots){
+ if(slot.getAnchor().equals(variable)){
+ return slot.getWords();
+ }
}
}
return null;
Modified: trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/templator/Templator.java
===================================================================
--- trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/templator/Templator.java 2011-10-26 08:41:40 UTC (rev 3321)
+++ trunk/components-ext/src/main/java/org/dllearner/algorithm/tbsl/templator/Templator.java 2011-10-26 09:03:57 UTC (rev 3322)
@@ -16,10 +16,9 @@
import org.dllearner.algorithm.tbsl.ltag.parser.LTAG_Lexicon_Constructor;
import org.dllearner.algorithm.tbsl.ltag.parser.Parser;
import org.dllearner.algorithm.tbsl.ltag.parser.Preprocessor;
-import org.dllearner.algorithm.tbsl.nlp.ApachePartOfSpeechTagger;
-import org.dllearner.algorithm.tbsl.nlp.Lemmatizer;
import org.dllearner.algorithm.tbsl.nlp.LingPipeLemmatizer;
import org.dllearner.algorithm.tbsl.nlp.PartOfSpeechTagger;
+import org.dllearner.algorithm.tbsl.nlp.StanfordPartOfSpeechTagger;
import org.dllearner.algorithm.tbsl.nlp.WordNet;
import org.dllearner.algorithm.tbsl.sem.drs.DRS;
import org.dllearner.algorithm.tbsl.sem.drs.UDRS;
@@ -63,8 +62,8 @@
g = LTAG_Constructor.construct(grammarFiles);
-// tagger = new StanfordPartOfSpeechTagger();
- tagger = new ApachePartOfSpeechTagger();
+ tagger = new StanfordPartOfSpeechTagger();
+// tagger = new ApachePartOfSpeechTagger();
p = new Parser();
p.SHOW_GRAMMAR = true;
@@ -93,6 +92,7 @@
}
else {
tagged = s;
+ s = extractSentence(tagged);
}
String newtagged = pp.condenseNominals(pp.findNEs(tagged,s));
@@ -137,11 +137,11 @@
if (!containsModuloRenaming(drses,drs)) {
// // DEBUG
-// System.out.println(dude);
-// System.out.println(drs);
-// for (Slot sl : slots) {
-// System.out.println(sl.toString());
-// }
+ System.out.println(dude);
+ System.out.println(drs);
+ for (Slot sl : slots) {
+ System.out.println(sl.toString());
+ }
// //
drses.add(drs);
@@ -186,6 +186,9 @@
newwords.addAll(wordnet.getBestSynonyms(wordnetpos,att));
}
}
+ if(newwords.isEmpty()){
+
+ }
if (newwords.isEmpty()) {
newwords.add(slot.getWords().get(0));
}
@@ -250,5 +253,23 @@
}
return false;
}
+
+ private String extractSentence(String taggedSentence){
+ int pos = taggedSentence.indexOf("/");
+ while(pos != -1){
+ String first = taggedSentence.substring(0, pos);
+ int endPos = taggedSentence.substring(pos).indexOf(" ");
+ if(endPos == -1){
+ endPos = taggedSentence.substring(pos).length();
+ }
+ String rest = taggedSentence.substring(pos + endPos);
+
+ taggedSentence = first + rest;
+ pos = taggedSentence.indexOf("/");
+
+ }
+ return taggedSentence;
+
+ }
}
This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
|
