/**
* Reads the threshold from the configuration. We use a threshold to decide whether it's an
* entailment (above the threshold) or not (below). This is intended for illustration purposes
* only, as the similarity scores are not normally distributed.
*/
private void initializeThreshold(CommonConfig config)
throws ConfigurationException, ComponentException {
NameValueTable EDA = null;
try {
EDA = config.getSection(this.getClass().getName());
} catch (ConfigurationException e) {
throw new ConfigurationException(e.getMessage() + " No EDA section.");
}
threshold = EDA.getDouble("threshold");
}
public void testTesting_MultiTH_AND_Output(CommonConfig config) {
MaxEntClassificationEDA meceda = new MaxEntClassificationEDA();
BufferedWriter output = null;
try {
meceda.initialize(config);
// check the test data directory
meceda.initializeData(config, false);
output =
new BufferedWriter(
new OutputStreamWriter(
new FileOutputStream(
config.getConfigurationFileName().replace("configuration-file", "results")
+ "_Result.txt"),
"UTF-8"));
logger.info("build CASes for input sentence pairs:");
int correct = 0;
int sum = 0;
for (File file : (new File(meceda.getTestDIR())).listFiles()) {
// ignore all the non-xmi files
if (!file.getName().endsWith(".xmi")) {
continue;
}
JCas cas = PlatformCASProber.probeXmi(file, null);
ClassificationTEDecision decision = meceda.process(cas);
output.write(decision.getPairID());
output.write("\t");
output.write(meceda.getGoldLabel(cas).toUpperCase());
output.write("\t");
output.write(decision.getDecision().toString().toUpperCase());
output.write("\t");
output.write(String.valueOf(decision.getConfidence()));
output.newLine();
logger.info("Pair " + decision.getPairID() + " is done.");
if (meceda.getGoldLabel(cas).equalsIgnoreCase(decision.getDecision().toString())) {
correct++;
}
sum++;
}
output.close();
logger.info("The correctly predicted pairs are " + correct + " / " + sum);
meceda.shutdown();
logger.info("EDA shuts down.");
} catch (Exception e) {
logger.info(e.getMessage());
}
}
/**
* the constructor using <code>CommonConfig</code>
*
* @param config the configuration
* @throws ConfigurationException
* @throws LexicalResourceException
*/
public BagOfLexesScoringEN(CommonConfig config)
throws ConfigurationException, LexicalResourceException {
NameValueTable comp = config.getSection("BagOfLexesScoring");
if (null == comp.getString("WordnetLexicalResource")
&& null == comp.getString("VerbOceanLexicalResource")) {
throw new ConfigurationException(
"Wrong configuation: didn't find any lexical resources for the BagOfLexesScoring component");
}
if (null != comp.getString("WordnetLexicalResource")) {
String[] WNRelations = comp.getString("WordnetLexicalResource").split(",");
if (null == WNRelations || 0 == WNRelations.length) {
throw new ConfigurationException(
"Wrong configuation: didn't find any relations for the WordNet");
}
Set<WordNetRelation> wnRelSet = new HashSet<WordNetRelation>();
for (String relation : WNRelations) {
if (relation.equalsIgnoreCase("HYPERNYM")) {
wnRelSet.add(WordNetRelation.HYPERNYM);
} else if (relation.equalsIgnoreCase("SYNONYM")) {
wnRelSet.add(WordNetRelation.SYNONYM);
} else if (relation.equalsIgnoreCase("PART_HOLONYM")) {
wnRelSet.add(WordNetRelation.PART_HOLONYM);
} else {
logger.warning("Warning: wrong relation names for the WordNet");
}
}
if (wnRelSet.isEmpty()) {
throw new ConfigurationException(
"Wrong configuation: didn't find any (correct) relations for the WordNet");
}
boolean isCollapsed = true;
boolean useFirstSenseOnlyLeft = false;
boolean useFirstSenseOnlyRight = false;
String wnPath = "/ontologies/EnglishWordNet-dict/";
NameValueTable wnComp = config.getSection("WordnetLexicalResource");
if (null != wnComp) {
if (null != wnComp.getString("isCollapsed")
&& !Boolean.parseBoolean(wnComp.getString("isCollapsed"))) {
isCollapsed = false;
}
if (null != wnComp.getString("useFirstSenseOnlyLeft")
&& Boolean.parseBoolean(wnComp.getString("useFirstSenseOnlyLeft"))) {
useFirstSenseOnlyLeft = true;
}
if (null != wnComp.getString("useFirstSenseOnlyRight")
&& Boolean.parseBoolean(wnComp.getString("useFirstSenseOnlyRight"))) {
useFirstSenseOnlyRight = true;
}
if (null != wnComp.getString("wordNetFilesPath")) {
wnPath = wnComp.getString("wordNetFilesPath");
}
}
wnlrSet = new HashSet<WordnetLexicalResource>();
File wnFile = new File(wnPath);
if (!wnFile.exists()) {
throw new ConfigurationException("cannot find WordNet at: " + wnPath);
}
if (isCollapsed) {
WordnetLexicalResource wnlr =
new WordnetLexicalResource(
wnFile, useFirstSenseOnlyLeft, useFirstSenseOnlyRight, wnRelSet);
wnlrSet.add(wnlr);
numOfFeats++;
} else {
for (WordNetRelation wnr : wnRelSet) {
WordnetLexicalResource wnlr =
new WordnetLexicalResource(
wnFile,
useFirstSenseOnlyLeft,
useFirstSenseOnlyRight,
Collections.singleton(wnr));
wnlrSet.add(wnlr);
numOfFeats++;
}
}
logger.info("Load WordNet done.");
}
if (null != comp.getString("VerbOceanLexicalResource")) {
String[] VORelations = comp.getString("VerbOceanLexicalResource").split(",");
if (null == VORelations || 0 == VORelations.length) {
throw new ConfigurationException(
"Wrong configuation: didn't find any relations for the VerbOcean");
}
Set<RelationType> voRelSet = new HashSet<RelationType>();
for (String relation : VORelations) {
if (relation.equalsIgnoreCase("strongerthan")) {
voRelSet.add(RelationType.STRONGER_THAN);
} else if (relation.equalsIgnoreCase("canresultin")) {
voRelSet.add(RelationType.CAN_RESULT_IN);
} else if (relation.equalsIgnoreCase("similar")) {
voRelSet.add(RelationType.SIMILAR);
} else {
logger.warning("Warning: wrong relation names for the VerbOcean");
}
}
if (voRelSet.isEmpty()) {
throw new ConfigurationException(
"Wrong configuation: didn't find any (correct) relations for the VerbOcean");
}
boolean isCollapsed = true;
String voPath = "/VerbOcean/verbocean.unrefined.2004-05-20.txt";
NameValueTable voComp = config.getSection("VerbOceanLexicalResource");
if (null != voComp) {
if (null != voComp.getString("isCollapsed")
&& !Boolean.parseBoolean(voComp.getString("isCollapsed"))) {
isCollapsed = false;
}
if (null != voComp.getString("verbOceanFilePath")) {
voPath = voComp.getString("verbOceanFilePath");
}
}
volrSet = new HashSet<VerbOceanLexicalResource>();
File voFile = new File(voPath);
if (!voFile.exists()) {
throw new ConfigurationException("cannot find VerbOcean at: " + voPath);
}
if (isCollapsed) {
VerbOceanLexicalResource volr = new VerbOceanLexicalResource(1, voFile, voRelSet);
volrSet.add(volr);
numOfFeats++;
} else {
for (RelationType vor : voRelSet) {
VerbOceanLexicalResource volr =
new VerbOceanLexicalResource(1, voFile, Collections.singleton(vor));
volrSet.add(volr);
numOfFeats++;
}
}
logger.info("Load VerbOcean done.");
}
}
/**
* Call this method once before starting to align sentence pairs.
*
* @param config a CommonConfig instance. The aligner retrieves the lexical resources
* configuration values.
* @throws LexicalResourceException if initialization of a resource failed
* @throws ConfigurationException if the configuration is invalid
*/
private void init(CommonConfig config) throws LexicalResourceException, ConfigurationException {
// Get the general parameters configuration section
NameValueTable paramsSection = null;
try {
paramsSection = config.getSection(GENERAL_PARAMS_CONF_SECTION);
} catch (ConfigurationException e) {
throw new ConfigurationException(e);
}
maxPhrase = paramsSection.getInteger(MAX_PHRASE_KEY);
// Get the Lexical Resources configuration section
NameValueTable lexicalResourcesSection = null;
try {
lexicalResourcesSection = config.getSection(LEXICAL_RESOURCES_CONF_SECTION);
} catch (ConfigurationException e) {
throw new ConfigurationException(e);
}
lexicalResources = new ArrayList<LexicalResource<? extends RuleInfo>>();
ConfigurationFile configFile = new ConfigurationFile(config);
// Get each resource and create it using the configuration section related to it
for (String resourceName : lexicalResourcesSection.keySet()) {
// Get the class name
String resourceClassName = lexicalResourcesSection.getString(resourceName);
// Get the configuration params
ConfigurationParams resourceParams = configFile.getModuleConfiguration(resourceName);
resourceParams.setExpandingEnvironmentVariables(true);
LexicalResource<? extends RuleInfo> lexicalResource =
createLexicalResource(resourceClassName, resourceParams);
if (lexicalResource != null) {
lexicalResources.add(lexicalResource);
PartOfSpeech leftSidePOS = null, rightSidePOS = null;
// Add the information about this resource
// Get the right and left side POS, in case it's mentioned
if (resourceParams.keySet().contains(LEFT_SIDE_POS_PARAM)) {
try {
leftSidePOS =
new ByCanonicalPartOfSpeech(resourceParams.getString(LEFT_SIDE_POS_PARAM));
} catch (UnsupportedPosTagStringException e) {
logger.warn(
"Could not load POS for left side: "
+ resourceParams.getString(LEFT_SIDE_POS_PARAM)
+ ". Alignment links of all POS will be retreived.");
}
}
if (resourceParams.keySet().contains(RIGHT_SIDE_POS_PARAM)) {
try {
rightSidePOS =
new ByCanonicalPartOfSpeech(resourceParams.getString(RIGHT_SIDE_POS_PARAM));
} catch (UnsupportedPosTagStringException e) {
logger.warn(
"Could not load POS for right side: "
+ resourceParams.getString(RIGHT_SIDE_POS_PARAM)
+ ". Alignment links of all POS will be retreived.");
}
}
lexicalResourcesInformation.put(
lexicalResource.getClass().getName(),
new LexicalResourceInformation(
resourceParams.getString(VERSION_PARAM),
resourceParams.getBoolean(USE_LEMMA_PARAM),
leftSidePOS,
rightSidePOS));
}
}
}