com.articulate.sigma.WordNet Maven / Gradle / Ivy
Show all versions of sigma-component Show documentation
/** This code is copyright Articulate Software (c) 2003-2007. Some portions
copyright Teknowledge (c) 2003 and reused under the terms of the GNU license.
This software is released under the GNU Public License .
Users of this code also consent, by use of this code, to credit Articulate Software
and Teknowledge in any writings, briefings, publications, presentations, or
other representations of any software which incorporates, builds on, or uses this
code. Please cite the following article in any publication with references:
Pease, A., (2003). The Sigma Ontology Development Environment,
in Working Notes of the IJCAI-2003 Workshop on Ontology and Distributed Systems,
August 9, Acapulco, Mexico. See also http://sigmakee.sourceforge.net
*/
package com.articulate.sigma;
import com.articulate.sigma.wordNet.MultiWords;
import java.io.BufferedReader;
import java.io.DataInputStream;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.FileReader;
import java.io.FileWriter;
import java.io.IOException;
import java.io.InputStreamReader;
import java.io.LineNumberReader;
import java.io.PrintWriter;
import java.util.*;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/** ***************************************************************
* This program finds and displays SUMO terms that are related in meaning to the English
* expressions that are entered as input. Note that this program uses four WordNet data
* files, "NOUN.EXC", "VERB.EXC" etc, as well as four WordNet to SUMO
* mappings files called "WordNetMappings-nouns.txt", "WordNetMappings-verbs.txt" etc
* The main part of the program prompts the user for an English term and then
* returns associated SUMO concepts. The two primary public methods are initOnce() and page().
* @author Ian Niles
* @author Adam Pease
*/
public class WordNet {
public static WordNet wn = new WordNet();
/* A map of language name to wordnets */
public static HashMap wns = new HashMap();
private static String baseDir = "";
private static File baseDirFile = null;
public static boolean initNeeded = true;
private static final String[][] wnFilenamePatterns =
{ { "noun_mappings", "WordNetMappings.*noun.*txt" },
{ "verb_mappings", "WordNetMappings.*verb.*txt" },
{ "adj_mappings", "WordNetMappings.*adj.*txt" },
{ "adv_mappings", "WordNetMappings.*adv.*txt" },
{ "noun_exceptions", "noun.exc" },
{ "verb_exceptions", "verb.exc" },
{ "adj_exceptions", "adj.exc" },
{ "adv_exceptions", "adv.exc" },
{ "sense_indexes", "index.sense" },
{ "word_frequencies", "wordFrequencies.txt" },
{ "cntlist", "cntlist" },
{ "stopwords", "stopwords.txt" },
{ "messages", "messages.txt" }
};
/** This array contains all of the compiled Pattern objects that
* will be used by methods in this file. */
private static Pattern[] regexPatterns = null;
public Hashtable nounSynsetHash = new Hashtable(); // Words in root form are String keys,
public Hashtable verbSynsetHash = new Hashtable(); // String values are synset lists.
public Hashtable adjectiveSynsetHash = new Hashtable();
public Hashtable adverbSynsetHash = new Hashtable();
public Hashtable verbDocumentationHash = new Hashtable(); // Keys are synset Strings, values
public Hashtable adjectiveDocumentationHash = new Hashtable(); // are documentation strings.
public Hashtable adverbDocumentationHash = new Hashtable();
public Hashtable nounDocumentationHash = new Hashtable();
public Hashtable nounSUMOHash = new Hashtable(); // Keys are synset Strings, values are SUMO
public Hashtable verbSUMOHash = new Hashtable(); // terms with the &% prefix and =, +, @ or [ suffix.
public Hashtable adjectiveSUMOHash = new Hashtable();
public Hashtable adverbSUMOHash = new Hashtable();
public String maxNounSynsetID = "";
public String maxVerbSynsetID = "";
/** Keys are SUMO terms, values are ArrayLists(s) of
* POS-prefixed synset String(s) meaning that the part of speech code is
* prepended to the synset number. */
public Hashtable> SUMOHash = new Hashtable>();
/** Keys are String POS-prefixed synsets. Values
* are ArrayList(s) of String(s) which are words. Note
* that the order of words in the file is preserved. */
public Hashtable> synsetsToWords = new Hashtable>();
/** list of irregular plural forms where the key is the
* plural, singular is the value. */
public Hashtable exceptionNounHash = new Hashtable();
// key is past tense, value is infinitive (without "to")
public Hashtable exceptionVerbHash = new Hashtable();
// The reverse index of the above
private Hashtable exceptionNounPluralHash = new Hashtable();
private Hashtable exceptionVerbPastHash = new Hashtable();
/** Keys are POS-prefixed synsets, values are ArrayList(s) of AVPair(s)
* in which the attribute is a pointer type according to
* http://wordnet.princeton.edu/man/wninput.5WN.html#sect3 and
* the value is a POS-prefixed synset @see WordNetUtilities.convertWordNetPointer */
public Hashtable> relations = new Hashtable>();
/** a HashMap of HashMaps where the key is a word sense of the
* form word_POS_num signifying the word, part of speech and number
* of the sense in WordNet. The value is a HashMap of words and the
* number of times that word cooccurs in sentences with the word sense
* given in the key. */
protected HashMap> wordCoFrequencies = new HashMap>();
/** a HashMap of HashMaps where the key is a word and the value is a
* HashMap of 9-digit POS-prefixed senses and the number of times that
* sense occurs in the Brown corpus. */
protected HashMap> wordFrequencies = new HashMap>();
/** a HashMap where the key is a 9-digit POS-prefixed sense and the value is a
* the number of times that sense occurs in the Brown corpus. */
public HashMap senseFrequencies = new HashMap();
/** English "stop words" such as "a", "at", "them", which have no or little
* inherent meaning when taken alone. */
public ArrayList stopwords = new ArrayList();
/** A HashMap where the keys are of the form word_POS_sensenum (alpha POS like "VB")
* and values are 8 digit WordNet synset byte offsets. */
public HashMap senseIndex = new HashMap();
/** A HashMap where the keys are 9 digit POS prefixed WordNet synset byte offsets,
* and the values are of the form word_POS_sensenum (alpha POS like "VB"). */
public HashMap reverseSenseIndex = new HashMap();
/** A HashMap where keys are 8 digit
* WordNet synset byte offsets or synsets appended with a dash and a specific
* word such as "12345678-foo". Values are ArrayList(s) of String
* verb frame numbers. */
public HashMap> verbFrames = new HashMap>();
/** A HashMap with words as keys and ArrayList as values. The
* ArrayList contains word senses which are Strings of the form
* word_POS_num (alpha POS like "VB") signifying the word, part of speech and number of
* the sense in WordNet. */
public HashMap> wordsToSenses = new HashMap>();
public MultiWords multiWords = new MultiWords();
//private Pattern p;
private Matcher m;
public static final int NOUN = 1;
public static final int VERB = 2;
public static final int ADJECTIVE = 3;
public static final int ADVERB = 4;
public static final int ADJECTIVE_SATELLITE = 5;
/** A HashMap with language name keys and HashMap values. The interior HashMap
* has String keys which are PWN30 synsets with 8-digit synsets a dash and then a alphabetic
* part of speech character. Values are words in the target language. */
public HashMap> OMW = new HashMap>();
/** This array contains all of the regular expression strings that
* will be compiled to Pattern objects for use in the methods in
* this file. */
private static final String[] regexPatternStrings =
{
// 0: WordNet.processPointers()
"^\\s*\\d\\d\\s\\S\\s\\d\\S\\s",
// 1: WordNet.processPointers()
"^([a-zA-Z0-9'._\\-]\\S*)\\s([0-9a-f])\\s",
// 2: WordNet.processPointers()
"^...\\s",
// 3: WordNet.processPointers()
"^(\\S\\S?)\\s([0-9]{8})\\s(.)\\s([0-9a-f]{4})\\s?",
// 4: WordNet.processPointers()
"^..\\s",
// 5: WordNet.processPointers()
"^\\+\\s(\\d\\d)\\s(\\d\\d)\\s?",
// 6: WordNet.readNouns()
"^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+?)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$",
// 7: WordNet.readNouns()
"^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+)$",
// 8: WordNet.readNouns()
"(\\S+)\\s+(\\S+)",
// 9: WordNet.readNouns()
"(\\S+)\\s+(\\S+)\\s+(\\S+)",
// 10: WordNet.readVerbs()
"^([0-9]{8})([^\\|]+)\\|\\s([\\S\\s]+?)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$",
// 11: WordNet.readVerbs()
"^([0-9]{8})([^\\|]+)\\|\\s([\\S\\s]+)$",
// 12: WordNet.readVerbs()
"(\\S+)\\s+(\\S+).*",
// 13: WordNet.readAdjectives()
"^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+?)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$",
// 14: WordNet.readAdjectives()
"^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+)$",
// 15: WordNet.readAdverbs()
"^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$",
// 16: WordNet.readAdverbs()
"^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+)$",
// 17: WordNet.readWordFrequencies()
"^Word: ([^ ]+) Values: (.*)",
// 18: WordNet.readSenseIndex()
"([^%]+)%([^:]*):[^:]*:[^:]*:[^:]*:[^ ]* ([^ ]+) ([^ ]+) .*",
// 19: WordNet.removePunctuation()
"(\\w)\\'re",
// 20: WordNet.removePunctuation()
"(\\w)\\'m",
// 21: WordNet.removePunctuation()
"(\\w)n\\'t",
// 22: WordNet.removePunctuation()
"(\\w)\\'ll",
// 23: WordNet.removePunctuation()
"(\\w)\\'s",
// 24: WordNet.removePunctuation()
"(\\w)\\'d",
// 25: WordNet.removePunctuation()
"(\\w)\\'ve"
};
public MultiWords getMultiWords() {
return multiWords;
}
/** ***************************************************************
* This method compiles all of the regular expression pattern
* strings in regexPatternStrings and puts the resulting compiled
* Pattern objects in the Pattern[] regexPatterns.
*/
private void compileRegexPatterns() {
System.out.println("INFO in WordNet.compileRegexPatterns(): compiling patterns");
regexPatterns = new Pattern[regexPatternStrings.length];
for (int i = 0; i < regexPatternStrings.length; i++) {
regexPatterns[i] = Pattern.compile(regexPatternStrings[i]);
if (!(regexPatterns[i] instanceof Pattern))
System.out.println("ERROR in WordNet.compileRegexPatterns(): could not compile \""
+ regexPatternStrings[i] + "\"");
}
return;
}
/** ***************************************************************
* Returns the WordNet File object corresponding to key. The
* purpose of this accessor is to make it easier to deal with
* possible changes to these file names, since the descriptive
* key, ideally, need not change. Each key maps to a regular
* expression that is used to match against filenames found in the
* directory denoted by WordNet.baseDir. If multiple filenames
* match the pattern for one key, then the file that was most
* recently changed (presumably, saved) is chosen.
*
* @param key A descriptive literal String that maps to a regular
* expression pattern used to obtain a WordNet file.
*
* @return A File object
*/
public File getWnFile(String key) {
File theFile = null;
try {
String pattern = null;
int i = 0;
for (i = 0; i < wnFilenamePatterns.length; i++) {
if ((wnFilenamePatterns[i][0]).equalsIgnoreCase(key)) {
pattern = wnFilenamePatterns[i][1];
break;
}
}
if ((pattern != null) && (baseDirFile != null)) {
File[] wnFiles = baseDirFile.listFiles();
if (wnFiles != null) {
for (i = 0; i < wnFiles.length; i++) {
if (wnFiles[i].getName().matches(pattern) && wnFiles[i].exists()) {
if (theFile != null) {
if (wnFiles[i].lastModified() > theFile.lastModified())
theFile = wnFiles[i];
}
else
theFile = wnFiles[i];
}
}
}
}
}
catch (Exception ex) {
System.out.println("Error in WordNet.getWnFile(): key: " + key);
System.out.println(ex.getMessage());
ex.printStackTrace();
}
return theFile;
}
/** ***************************************************************
* Add a synset (with part of speech number prefix) and the SUMO
* term that maps to it.
*/
private void addSUMOHash(String term, String synset) {
//System.out.println("INFO in WordNet.addSUMOHash(): SUMO term: " + key);
//System.out.println("INFO in WordNet.addSUMOHash(): synset: " + value);
term = term.substring(2,term.length()-1);
ArrayList synsets = SUMOHash.get(term);
if (synsets == null) {
synsets = new ArrayList();
SUMOHash.put(term,synsets);
}
synsets.add(synset);
}
/** ***************************************************************
* Return an ArrayList of the string split by spaces.
*/
public static ArrayList splitToArrayList(String st) {
if (StringUtil.emptyString(st)) {
System.out.println("Error in WordNet.splitToArrayList(): empty string input");
return null;
}
String[] sentar = st.split(" ");
ArrayList words = new ArrayList(Arrays.asList(sentar));
for (int i = 0; i < words.size(); i++) {
if (words.get(i).equals("") || words.get(i) == null || words.get(i).matches("\\s*"))
words.remove(i);
}
return words;
}
/** ***************************************************************
* Return an ArrayList of the string split by periods.
*/
public static ArrayList splitToArrayListSentence(String st) {
if (st.equals("") || st == null) {
System.out.println("Error in WordNet.splitToArrayList(): empty string input");
return null;
}
String[] sentar = st.split("\\.\\s");
ArrayList words = new ArrayList(Arrays.asList(sentar));
for (int i = 0; i < words.size(); i++) {
if (words.get(i).equals("") || words.get(i) == null || words.get(i).matches("\\s*"))
words.remove(i);
}
return words;
}
/** ***************************************************************
* Add a synset and its corresponding word to the synsetsToWords
* variable. Prefix the synset with its part of speech before adding.
*/
private void addToSynsetsToWords(String word, String synsetStr, String POS) {
if (word.indexOf('_') >= 0)
multiWords.addMultiWord(word);
ArrayList al = synsetsToWords.get(POS + synsetStr);
if (al == null) {
al = new ArrayList();
synsetsToWords.put(POS + synsetStr, al);
}
al.add(word);
switch (POS.charAt(0)) {
case '1':
String synsets = (String) nounSynsetHash.get(word);
if (synsets == null)
synsets = "";
if (synsets.indexOf(synsetStr) < 0) {
if (synsets.length() > 0)
synsets = synsets + " ";
synsets = synsets + synsetStr;
nounSynsetHash.put(word,synsets);
}
break;
case '2':
synsets = (String) verbSynsetHash.get(word);
if (synsets == null)
synsets = "";
if (synsets.indexOf(synsetStr) < 0) {
if (synsets.length() > 0)
synsets = synsets + " ";
synsets = synsets + synsetStr;
//if (word.equals("roll"))
// System.out.println("INFO in WordNet.addToSynsetsToWords(): word: " + word);
verbSynsetHash.put(word,synsets);
}
break;
case '3':
synsets = (String) adjectiveSynsetHash.get(word);
if (synsets == null)
synsets = "";
if (synsets.indexOf(synsetStr) < 0) {
if (synsets.length() > 0)
synsets = synsets + " ";
synsets = synsets + synsetStr;
adjectiveSynsetHash.put(word,synsets);
}
break;
case '4':
synsets = (String) adverbSynsetHash.get(word);
if (synsets == null)
synsets = "";
if (synsets.indexOf(synsetStr) < 0) {
if (synsets.length() > 0)
synsets = synsets + " ";
synsets = synsets + synsetStr;
adverbSynsetHash.put(word,synsets);
}
break;
}
}
/** ***************************************************************
* Process some of the fields in a WordNet .DAT file as described at
* http://wordnet.princeton.edu/man/wndb.5WN . synset must include
* the POS-prefix. Input should be of the form
* lex_filenum ss_type w_cnt word lex_id [word lex_id...] p_cnt [ptr...] [frames...]
*/
private void processPointers(String synset, String pointers) {
//System.out.println("INFO in WordNet.processPointers(): " + pointers);
// 0: p = Pattern.compile("^\\s*\\d\\d\\s\\S\\s\\d\\S\\s");
m = regexPatterns[0].matcher(pointers);
pointers = m.replaceFirst("");
//System.out.println("INFO in WordNet.processPointers(): removed prefix: " + pointers);
// Should be left with:
// word lex_id [word lex_id...] p_cnt [ptr...] [frames...]
// 1: p = Pattern.compile("^([a-zA-Z0-9'._\\-]\\S*)\\s([0-9a-f])\\s");
m = regexPatterns[1].matcher(pointers);
while (m.lookingAt()) {
String word = m.group(1);
//if (word.equals("roll"))
// System.out.println("INFO in WordNet.processPointers(): word: " + word);
if (word.length() > 3 && (word.substring(word.length()-3,word.length()).equals("(a)") ||
word.substring(word.length()-3,word.length()).equals("(p)")))
word = word.substring(0,word.length()-3);
if (word.length() > 4 && word.substring(word.length()-4,word.length()).equals("(ip)"))
word = word.substring(0,word.length()-4);
//String count = m.group(2);
addToSynsetsToWords(word,synset.substring(1),synset.substring(0,1));
pointers = m.replaceFirst("");
m = regexPatterns[1].matcher(pointers);
}
//System.out.println("INFO in WordNet.processPointers(): removed words: " + pointers);
// Should be left with:
// p_cnt [ptr...] [frames...]
// 2: p = Pattern.compile("^...\\s");
m = regexPatterns[2].matcher(pointers);
pointers = m.replaceFirst("");
// Should be left with:
// [ptr...] [frames...]
// where ptr is
// pointer_symbol synset_offset pos source/target
// 3: p = Pattern.compile("^(\\S\\S?)\\s([0-9]{8})\\s(.)\\s([0-9a-f]{4})\\s?");
m = regexPatterns[3].matcher(pointers);
while (m.lookingAt()) {
String ptr = m.group(1);
String targetSynset = m.group(2);
String targetPOS = m.group(3);
//String sourceTarget = m.group(4);
targetPOS = (new Character(WordNetUtilities.posLetterToNumber(targetPOS.charAt(0)))).toString();
pointers = m.replaceFirst("");
m = regexPatterns[3].matcher(pointers);
ptr = WordNetUtilities.convertWordNetPointer(ptr);
AVPair avp = new AVPair();
avp.attribute = ptr;
avp.value = targetPOS + targetSynset;
ArrayList al = new ArrayList();
if (relations.keySet().contains(synset))
al = relations.get(synset);
else {
relations.put(synset,al);
}
//System.out.println("INFO in WordNet.processPointers(): (" + avp.attribute +
// " " + synset + " " + avp.value);
al.add(avp);
}
if ( (pointers != null)
&& (pointers != "")
&& (pointers.length() > 0)
&& !pointers.equals(" ") ) {
// Only for verbs may we have the following leftover
// f_cnt + f_num w_num [ + f_num w_num...]
if (synset.charAt(0) == '2') {
// 4: p = Pattern.compile("^..\\s");
m = regexPatterns[4].matcher(pointers);
pointers = m.replaceFirst("");
// 5: p = Pattern.compile("^\\+\\s(\\d\\d)\\s(\\d\\d)\\s?");
m = regexPatterns[5].matcher(pointers);
while (m.lookingAt()) {
String frameNum = m.group(1);
String wordNum = m.group(2);
String key;
if (wordNum.equals("00"))
key = synset.substring(1);
else {
int num = Integer.valueOf(wordNum).intValue();
ArrayList al = synsetsToWords.get(synset);
if (al == null)
System.out.println("Error in WordNet.processPointers(): " + synset
+ " has no words for pointers: \"" + pointers + "\"");
String word = (String) al.get(num-1);
key = synset.substring(1) + "-" + word;
}
ArrayList frames = new ArrayList();
if (!verbFrames.keySet().contains(key))
verbFrames.put(key,frames);
else
frames = verbFrames.get(key);
frames.add(frameNum);
pointers = m.replaceFirst("");
m = regexPatterns[5].matcher(pointers);
}
}
else {
System.out.println("Error in WordNet.processPointers(): " + synset.charAt(0) + " leftover pointers: \"" + pointers + "\"");
}
}
return;
}
/** ***************************************************************
*/
private void addSUMOMapping(String SUMO, String synset) {
SUMO = SUMO.trim();
switch (synset.charAt(0)) {
case '1': nounSUMOHash.put(synset.substring(1),SUMO);
break;
case '2': verbSUMOHash.put(synset.substring(1),SUMO);
break;
case '3': adjectiveSUMOHash.put(synset.substring(1),SUMO);
break;
case '4': adverbSUMOHash.put(synset.substring(1),SUMO);
break;
}
addSUMOHash(SUMO, synset);
}
/** ***************************************************************
* Get the SUMO mapping for a POS-prefixed synset
*/
public String getSUMOMapping(String synset) {
if (StringUtil.emptyString(synset)) {
System.out.println("Error in WordNet.getSUMOMapping: null synset ");
return null;
}
switch (synset.charAt(0)) {
case '1': return (String) nounSUMOHash.get(synset.substring(1));
case '2': return (String) verbSUMOHash.get(synset.substring(1));
case '3': return (String) adjectiveSUMOHash.get(synset.substring(1));
case '4': return (String) adverbSUMOHash.get(synset.substring(1));
}
System.out.println("Error in WordNet.getSUMOMapping: improper first character for synset: " + synset);
return null;
}
/** ***************************************************************
* Create the hashtables nounSynsetHash, nounDocumentationHash,
* nounSUMOhash and exceptionNounHash that contain the WordNet
* noun synsets, word definitions, mappings to SUMO, and plural
* exception forms, respectively.
* Throws an IOException if the files are not found.
*/
private void readNouns() throws java.io.IOException {
System.out.println("INFO in WordNet.readNouns(): Reading WordNet noun files");
LineNumberReader lr = null;
try {
// synset_offset lex_filenum ss_type w_cnt word lex_id [word lex_id...] p_cnt [ptr...] [frames...] | gloss
String line;
File nounFile = getWnFile("noun_mappings");
if (nounFile == null) {
System.out.println("Error in WordNet.readNouns(): The noun mappings file does not exist in " + baseDir );
return;
}
long t1 = System.currentTimeMillis();
FileReader r = new FileReader(nounFile);
// System.out.println( "INFO in WordNet.readNouns(): Reading file " + nounFile.getCanonicalPath() );
lr = new LineNumberReader(r);
while ((line = lr.readLine()) != null) {
if (lr.getLineNumber() % 1000 == 0)
System.out.print('.');
line = line.trim();
if(!processNounLine(line)) {
System.out.println();
System.out.println( "Error in WordNet.readNouns(): No match in "
+ nounFile.getCanonicalPath() + " for line " + line );
}
}
System.out.println("x");
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + nounFile.getCanonicalPath() +
" with " + lr.getLineNumber() + " lines");
// System.out.println("INFO in WordNet.readNouns(): Reading WordNet noun exceptions");
nounFile = getWnFile("noun_exceptions");
if (nounFile == null) {
System.out.println("ERROR in WordNet.readNouns(): "
+ "The noun mapping exceptions file does not exist in " + baseDir);
return;
}
t1 = System.currentTimeMillis();
r = new FileReader( nounFile );
lr = new LineNumberReader(r);
while ((line = lr.readLine()) != null) {
// 8: p = Pattern.compile("(\\S+)\\s+(\\S+)");
m = regexPatterns[8].matcher(line);
if (m.matches()) {
exceptionNounHash.put(m.group(1),m.group(2)); // 1-plural, 2-singular
exceptionNounPluralHash.put(m.group(2),m.group(1));
}
else {
// 9: p = Pattern.compile("(\\S+)\\s+(\\S+)\\s+(\\S+)");
m = regexPatterns[9].matcher(line);
if (m.matches()) {
exceptionNounHash.put(m.group(1),m.group(2)); // 1-plural, 2-singular 3-alternate singular
exceptionNounPluralHash.put(m.group(2),m.group(1));
exceptionNounPluralHash.put(m.group(3),m.group(1));
}
else
if (line != null && line.length() > 0 && line.charAt(0) != ';') {
System.out.println("Error in WordNet.readNouns(): No match in "
+ nounFile.getCanonicalPath() + " for line " + line );
}
}
}
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + nounFile.getCanonicalPath() );
}
catch (Exception ex) {
ex.printStackTrace();
}
finally {
try {
if (lr != null) {
lr.close();
}
}
catch (Exception ex) {
}
}
return;
}
/** ***************************************************************
*/
protected boolean processNounLine(String line) {
// 6: p = Pattern.compile("^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+?)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$");
m = regexPatterns[6].matcher(line);
boolean anyAreNull = false;
if (m.matches()) {
for ( int i = 1 ; i < 5 ; i++ ) {
anyAreNull = (m.group(i) == null);
if ( anyAreNull ) {
break;
}
}
if ( ! anyAreNull ) {
addSUMOMapping(m.group(4),"1" + m.group(1));
maxNounSynsetID = m.group(1);
nounDocumentationHash.put(m.group(1),m.group(3)); // 1-synset, 2-pointers, 3-docu, 4-SUMO term
processPointers("1" + m.group(1),m.group(2));
}
}
else {
// 7: p = Pattern.compile("^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+)$"); // no SUMO mapping
m = regexPatterns[7].matcher(line);
if (m.matches()) {
nounDocumentationHash.put(m.group(1),m.group(3));
maxNounSynsetID = m.group(1);
processPointers("1" + m.group(1),m.group(2));
}
else {
//System.out.println("line: " + line);
if (line.length() > 0 && line.charAt(0) != ';') {
return false;
}
}
}
return true;
}
/** ***************************************************************
* Create the hashtables verbSynsetHash (by calling processPointers which calls
* addSynsetsToWords), verbDocumentationHash,
* verbSUMOhash and exceptionVerbHash that contain the WordNet
* verb synsets, word definitions, mappings to SUMO, and plural
* exception forms, respectively.
* Throws an IOException if the files are not found.
*/
private void readVerbs() throws java.io.IOException {
System.out.println("INFO in WordNet.readVerbs(): Reading WordNet verb files");
LineNumberReader lr = null;
try {
String line;
File verbFile = getWnFile("verb_mappings");
if (verbFile == null) {
System.out.println("Error in WordNet.readVerbs(): The verb mappings file does not exist in " + baseDir);
return;
}
long t1 = System.currentTimeMillis();
FileReader r = new FileReader(verbFile);
lr = new LineNumberReader(r);
while ((line = lr.readLine()) != null) {
if (lr.getLineNumber() % 1000 == 0)
System.out.print('.');
line = line.trim();
// 10: p = Pattern.compile("^([0-9]{8})([^\\|]+)\\|\\s([\\S\\s]+?)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$");
m = regexPatterns[10].matcher(line);
if (m.matches()) {
verbDocumentationHash.put(m.group(1),m.group(3));
maxVerbSynsetID = m.group(1);
addSUMOMapping(m.group(4),"2" + m.group(1));
processPointers("2" + m.group(1),m.group(2));
}
else {
// 11: p = Pattern.compile("^([0-9]{8})([^\\|]+)\\|\\s([\\S\\s]+)$"); // no SUMO mapping
m = regexPatterns[11].matcher(line);
if (m.matches()) {
verbDocumentationHash.put(m.group(1),m.group(3));
maxVerbSynsetID = m.group(1);
processPointers("2" + m.group(1),m.group(2));
}
else {
//System.out.println("line: " + line);
if (line != null && line.length() > 0 && line.charAt(0) != ';') {
System.out.println();
System.out.println("Error in WordNet.readVerbs(): No match in "
+ verbFile.getCanonicalPath() + " for line " + line );
}
}
}
}
System.out.println("x");
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + verbFile.getCanonicalPath() +
" with " + lr.getLineNumber() + " lines");
// System.out.println("INFO in WordNet.readVerbs(): Reading WordNet verb exceptions");
verbFile = getWnFile("verb_exceptions");
if (verbFile == null) {
System.out.println("Error in WordNet.readVerbs(): The verb mapping exceptions file does not exist in " + baseDir);
return;
}
t1 = System.currentTimeMillis();
r = new FileReader( verbFile );
lr = new LineNumberReader(r);
while ((line = lr.readLine()) != null) {
// 12: p = Pattern.compile("(\\S+)\\s+(\\S+).*");
m = regexPatterns[12].matcher(line); // TODO: Note we ignore more then one base form
if (m.matches()) {
exceptionVerbHash.put(m.group(1),m.group(2)); // 1-past, 2-infinitive
exceptionVerbPastHash.put(m.group(2),m.group(1));
}
else
if (line != null && line.length() > 0 && line.charAt(0) != ';')
System.out.println( "Error in WordNet.readVerbs(): No match in "
+ verbFile.getCanonicalPath() + " for line " + line );
}
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + verbFile.getCanonicalPath() );
}
catch (Exception ex) {
ex.printStackTrace();
}
finally {
try {
if (lr != null) {
lr.close();
}
}
catch (Exception ex) {
}
}
return;
}
/** ***************************************************************
* Create the hashtables adjectiveSynsetHash, adjectiveDocumentationHash,
* and adjectiveSUMOhash that contain the WordNet
* adjective synsets, word definitions, and mappings to SUMO, respectively.
* Throws an IOException if the files are not found.
*/
private void readAdjectives() throws java.io.IOException {
System.out.println("INFO in WordNet.readAdjectives(): Reading WordNet adjective files");
LineNumberReader lr = null;
try {
String line;
File adjFile = getWnFile("adj_mappings");
if (adjFile == null) {
System.out.println("Error in WordNet.readAdjectives(): The adjective mappings file does not exist in " + baseDir);
return;
}
long t1 = System.currentTimeMillis();
FileReader r = new FileReader( adjFile );
lr = new LineNumberReader(r);
while ((line = lr.readLine()) != null) {
if (lr.getLineNumber() % 1000 == 0)
System.out.print('.');
line = line.trim();
// 13: p = Pattern.compile("^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+?)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$");
m = regexPatterns[13].matcher(line);
if (m.matches()) {
adjectiveDocumentationHash.put(m.group(1),m.group(3));
addSUMOMapping(m.group(4), "3" + m.group(1));
processPointers("3" + m.group(1), m.group(2));
}
else {
// 14: p = Pattern.compile("^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+)$"); // no SUMO mapping
m = regexPatterns[14].matcher(line);
if (m.matches()) {
adjectiveDocumentationHash.put(m.group(1),m.group(3));
processPointers("3" + m.group(1),m.group(2));
}
else {
//System.out.println("line: " + line);
if (line != null && line.length() > 0 && line.charAt(0) != ';') {
System.out.println();
System.out.println( "Error in WordNet.readAdjectives(): No match in "
+ adjFile.getCanonicalPath() + " for line " + line );
}
}
}
}
System.out.println("x");
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + adjFile.getCanonicalPath() +
" with " + lr.getLineNumber() + " lines");
}
catch (Exception ex) {
ex.printStackTrace();
}
finally {
try {
if (lr != null) {
lr.close();
}
}
catch (Exception ex) {
}
}
return;
}
/** ***************************************************************
* Create the hashtables adverbSynsetHash, adverbDocumentationHash,
* and adverbSUMOhash that contain the WordNet
* adverb synsets, word definitions, and mappings to SUMO, respectively.
* Throws an IOException if the files are not found.
*/
private void readAdverbs() throws java.io.IOException {
System.out.println("INFO in WordNet.readAdverbs(): Reading WordNet adverb files");
LineNumberReader lr = null;
try {
String line;
File advFile = getWnFile("adv_mappings");
if (advFile == null) {
System.out.println("Error in WordNet.readAdverbs(): The adverb mappings file does not exist in " + baseDir);
return;
}
long t1 = System.currentTimeMillis();
FileReader r = new FileReader( advFile );
lr = new LineNumberReader(r);
while ((line = lr.readLine()) != null) {
if (lr.getLineNumber() % 1000 == 0)
System.out.print('.');
line = line.trim();
// 15: p = Pattern.compile("^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$");
m = regexPatterns[15].matcher(line);
if (m.matches()) {
adverbDocumentationHash.put(m.group(1),m.group(3));
addSUMOMapping(m.group(4),"4" + m.group(1));
processPointers("4" + m.group(1),m.group(2));
}
else {
// 16: p = Pattern.compile("^([0-9]{8})([\\S\\s]+)\\|\\s([\\S\\s]+)$"); // no SUMO mapping
m = regexPatterns[16].matcher(line);
if (m.matches()) {
adverbDocumentationHash.put(m.group(1),m.group(3));
processPointers("4" + m.group(1),m.group(2));
}
else {
//System.out.println("line: " + line);
if (line != null && line.length() > 0 && line.charAt(0) != ';') {
System.out.println();
System.out.println("Error in WordNet.readAdverbs(): No match in "
+ advFile.getCanonicalPath() + " for line " + line);
}
}
}
}
System.out.println("x");
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + advFile.getCanonicalPath() +
" with " + lr.getLineNumber() + " lines");
}
catch (Exception ex) {
ex.printStackTrace();
}
finally {
try {
if (lr != null) {
lr.close();
}
}
catch (Exception ex) {
}
}
return;
}
/** ***************************************************************
* Return a HashMap of HashMaps where the key is a word sense of the
* form word_POS_num signifying the word, part of speech and number
* of the sense in WordNet. The value is a HashMap of words and the
* number of times that word cooccurs in sentences with the word sense
* given in the key.
*/
public void readWordCoFrequencies() {
System.out.println("INFO in WordNet.readWordFrequencies(): Reading WordNet word frequencies");
LineNumberReader lr = null;
int counter = 0;
File wfFile = null;
String canonicalPath = "";
try {
wfFile = getWnFile("word_frequencies");
if (wfFile == null) {
System.out.println("Error in WordNet.readWordFrequencies(): The word frequencies file does not exist in " + baseDir);
return;
}
canonicalPath = wfFile.getCanonicalPath();
long t1 = System.currentTimeMillis();
FileReader r = new FileReader(wfFile);
lr = new LineNumberReader(r);
String line = null;
while ((line = lr.readLine()) != null) {
line = line.trim();
// 17: Pattern p = Pattern.compile("^Word: ([^ ]+) Values: (.*)");
Matcher m = regexPatterns[17].matcher(line);
if (m.matches()) {
String key = m.group(1);
String values = m.group(2);
String[] words = values.split(" ");
HashMap frequencies = new HashMap();
for (int i = 0; i < words.length-3; i++) {
if (words[i].equals("SUMOterm:")) {
i = words.length;
}
else {
if (words[i].indexOf("_") == -1) {
//System.out.println("INFO in WordNet.readWordFrequencies(). word: " + words[i]);
//System.out.println("INFO in WordNet.readWordFrequencies(). line: " + line);
}
else {
String word = words[i].substring(0,words[i].indexOf("_"));
String freq = words[i].substring(words[i].lastIndexOf("_") + 1, words[i].length());
frequencies.put(word.intern(),Integer.decode(freq));
}
}
}
wordCoFrequencies.put(key.intern(),frequencies);
counter++;
if (counter == 1000) {
System.out.print(".");
counter = 0;
}
}
}
System.out.println("x");
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + canonicalPath);
}
catch (Exception i) {
System.out.println();
System.out.println("Error in WordNet.readWordFrequencies() reading file "
+ canonicalPath + ": " + i.getMessage());
i.printStackTrace();
}
finally {
try {
if (lr != null) {
lr.close();
}
}
catch (Exception ex) {
}
}
return;
}
/** ***************************************************************
*/
public void readStopWords() {
System.out.println("INFO in WordNet.readStopWords(): Reading stop words");
LineNumberReader lr = null;
File swFile = null;
String canonicalPath = "";
try {
swFile = getWnFile("stopwords");
if (swFile == null) {
System.out.println("Error in WordNet.readStopWords(): The stopwords file does not exist in " + baseDir);
return;
}
canonicalPath = swFile.getCanonicalPath();
long t1 = System.currentTimeMillis();
FileReader r = new FileReader(swFile);
lr = new LineNumberReader(r);
String line;
while ((line = lr.readLine()) != null)
stopwords.add(line.intern());
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + canonicalPath );
}
catch (Exception i) {
System.out.println("Error in WordNet.readStopWords() reading file "
+ canonicalPath + ": " + i.getMessage());
i.printStackTrace();
}
finally {
try {
if (lr != null) {
lr.close();
}
}
catch (Exception ex) {
}
}
return;
}
/** ***************************************************************
* Note that WordNet forces all these words to lowercase in the index.xxx files
*/
public void readSenseIndex() {
System.out.println("INFO in WordNet.readSenseIndex(): Reading WordNet sense index");
LineNumberReader lr = null;
int counter = 0;
int totalcount = 0;
File siFile = null;
String canonicalPath = "";
try {
siFile = getWnFile("sense_indexes");
if (siFile == null) {
System.out.println("Error in WordNet.readSenseIndex(): The sense indexes file does not exist in " + baseDir);
return;
}
canonicalPath = siFile.getCanonicalPath();
long t1 = System.currentTimeMillis();
FileReader r = new FileReader( siFile );
lr = new LineNumberReader(r);
//System.out.println("INFO in WordNet.readSenseIndex(). Opened file.");
Matcher m = null;
String line;
while ((line = lr.readLine()) != null) {
// 18: Pattern p = Pattern.compile("([^%]+)%([^:]*):[^:]*:[^:]*:[^:]*:[^ ]* ([^ ]+) ([^ ]+) .*");
m = regexPatterns[18].matcher(line);
if (m.matches()) {
String word = m.group(1);
String pos = m.group(2);
String synset = m.group(3);
String sensenum = m.group(4);
String posString = WordNetUtilities.posNumberToLetters(pos); // alpha POS - NN,VB etc
String key = word + "_" + posString + "_" + sensenum;
ArrayList al = wordsToSenses.get(word);
if (al == null) {
al = new ArrayList();
wordsToSenses.put(word, al);
}
al.add(key);
senseIndex.put(key, synset);
reverseSenseIndex.put(pos + synset, key);
counter++;
if (counter == 1000) {
//System.out.println("INFO in WordNet.readSenseIndex(). Read word sense: " + key);
//System.out.println(word + " " + pos + " " + synset + " " + sensenum);
System.out.print('.');
totalcount = totalcount + counter;
counter = 0;
}
}
}
System.out.println("x");
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + canonicalPath + " with " + totalcount + " senses.");
}
catch (Exception i) {
System.out.println();
System.out.println( "Error in WordNet.readSenseIndex() reading file "
+ canonicalPath + ": " + i.getMessage());
i.printStackTrace();
}
finally {
try {
if (lr != null) {
lr.close();
}
}
catch (Exception ex) {
}
}
return;
}
/** ***************************************************************
* Read word sense frequencies into a HashMap of PriorityQueues
* containing AVPairs where the value is a word and the attribute
* (on which PriorityQueue is sorted) is an 8 digit String
* representation of an integer count.
*/
public void readSenseCount() {
System.out.println("INFO in WordNet.readSenseCount(): Reading WordNet sense counts");
LineNumberReader lr = null;
int counter = 0;
int totalcount = 0;
File siFile = null;
String canonicalPath = "";
try {
siFile = getWnFile("cntlist");
if (siFile == null) {
System.out.println("Error in WordNet.readSenseCount(): The sense count file does not exist in " +
baseDir + File.separator + "cntlist");
return;
}
canonicalPath = siFile.getCanonicalPath();
long t1 = System.currentTimeMillis();
FileReader r = new FileReader( siFile );
lr = new LineNumberReader(r);
//System.out.println("INFO in WordNet.readSenseIndex(). Opened file.");
Matcher m = null;
String line;
while ((line = lr.readLine()) != null) {
//System.out.println("Info in WordNet.readSenseCount(): line " + line);
Pattern p = Pattern.compile("([^ ]+) ([^%]+)%([^:]*):[^:]*:[^:]*:[^:]*:[^ ]* ([^ ]+)");
m = p.matcher(line);
if (m.matches()) {
String count = m.group(1);
String word = m.group(2);
String POS = m.group(3);
String posString = WordNetUtilities.posNumberToLetters(POS);
String sensenum = m.group(4);
String key = word + "_" + posString + "_" + sensenum; // word_POS_sensenum
AVPair avp = new AVPair();
avp.attribute = StringUtil.fillString(count, ' ', 8, true);
avp.value = key;
TreeSet pq = new TreeSet();
if (wordFrequencies.containsKey(word))
pq = wordFrequencies.get(word);
pq.add(avp);
//System.out.println("Info in WordNet.readSenseCount(): word, sense, count " + word +
// ", " + key + ", " + count);
wordFrequencies.put(word, pq);
int freq = Integer.parseInt(count);
String synset = WordNetUtilities.getSenseFromKey(key);
senseFrequencies.put(synset,freq);
//System.out.println(WordNet.wn.wordFrequencies.get("be"));
}
counter++;
if (counter == 1000) {
//System.out.println("INFO in WordNet.readSenseIndex(). Read word sense: " + key);
//System.out.println(word + " " + pos + " " + synset + " " + sensenum);
System.out.print('.');
totalcount = totalcount + counter;
counter = 0;
}
}
System.out.println("x");
System.out.println(" " + ((System.currentTimeMillis() - t1) / 1000.0)
+ " seconds to process " + canonicalPath + " with " + totalcount + " senses.");
}
catch (Exception i) {
System.out.println();
System.out.println( "Error in WordNet.readSenseCount() reading file "
+ canonicalPath + ": " + i.getMessage());
i.printStackTrace();
}
finally {
try {
if (lr != null) {
lr.close();
}
}
catch (Exception ex) {
}
}
return;
}
/** ***************************************************************
* A routine which looks up a given list of words in the hashtables
* to find the relevant word definitions and SUMO mappings.
* @param input is the target sentence to be parsed. See WordSenseBody.jsp for usage.
* @param context is the larger context of the sentence. Can mean more accurate results.
* @param params is the set of html parameters
* @returns a String that is the sentence taken apart and displayed in HTML
*/
public String sumoSentenceDisplay(String input, String context, String params) {
try {
ArrayList sentenceList = splitToArrayListSentence(input);
StringBuffer result = new StringBuffer();
for (String sentence : sentenceList) {
ArrayList synsetList = WSD.collectWordSenses(sentence); //returns an ArrayList of synsets
System.out.println("INFO in WordNet.sumoSentenceDisplay(): " + synsetList);
result.append("Target Sentence: " + sentence + "
\n");
//if (wordSynsetList == null || wordSynsetList.size() < 2)
// continue;
//String wordColl = wordSynsetList.get(0);
//String synsetColl = wordSynsetList.get(1);
//ArrayList synsetList = splitToArrayList(synsetColl); //removes all punctuation and whitespace
//ArrayList wordList = splitToArrayList(wordColl);
result.append(sumoSentimentDisplay(sentence)); //attaches sentiment display already in HTML format
HashMap conceptMap = (HashMap) DB.computeConceptSentiment(sentence); //returns Hashmap of key=SUMO concept, value=sentiment score
System.out.println("INFO in WordNet.sumoSentenceDisplay(): map " + conceptMap);
if (synsetList == null) {
result.append(" No words could be mapped for Sense Analysis. \n");
return result.toString();
}
ArrayList SUMOtermList = new ArrayList();
int listLength = synsetList.size();
for (int i = 0; i < listLength; i++) {
String s = synsetList.get(i);
String trm = getSUMOMapping(s); //returns a SUMO term in the form "&%term=" that matches the synset
SUMOtermList.add(trm);
}
String synset = new String();
String word = new String();
String SUMOterm = new String();
String documentation = new String();
for (int j = 0; j < listLength; j++) {
synset = synsetList.get(j);
ArrayList words = synsetsToWords.get(synset);
if (words != null && words.size() > 0)
word = words.get(0);
else
word = "";
SUMOterm = SUMOtermList.get(j);
if (nounDocumentationHash.containsKey(synset.substring(1)) && synset.substring(0,1).equals("1")) {
documentation = (String) nounDocumentationHash.get(synset.substring(1));
result.append("" + synset + " ");
result.append(" " + "" + word + "" + "- " + documentation + ".\n");
}
else if (verbDocumentationHash.containsKey(synset.substring(1)) && synset.substring(0,1).equals("2")) {
documentation = (String) verbDocumentationHash.get(synset.substring(1));
result.append("" + synset + " ");
result.append(" " + "" + word + "" + "- " + documentation + ".\n");
}
else if (adjectiveDocumentationHash.containsKey(synset.substring(1)) && synset.substring(0,1).equals("3")) {
documentation = (String) adjectiveDocumentationHash.get(synset.substring(1));
result.append("" + synset + " ");
result.append(" " + "" + word + "" + "- " + documentation + ".\n");
}
else if (adverbDocumentationHash.containsKey(synset.substring(1)) && synset.substring(0,1).equals("4")) {
documentation = (String) adverbDocumentationHash.get(synset.substring(1));
result.append("" + synset + " ");
result.append(" " + "1" + word + "" + "- " + documentation + ".\n");
}
if (SUMOterm == null)
result.append("" + word + " not yet mapped to SUMO
");
else {
result.append(HTMLformatter.termMappingsList(SUMOterm,"- \tSentiment Score in Context: " + conceptScore.toString() + "
");
}
}
}
result.append("
");
}
return result.toString();
}
catch (NullPointerException ex) {
System.out.println(ex.getMessage());
ex.printStackTrace();
return "Error";
}
}
/** ***************************************************************
* A routine that uses computeSentiment in DB.java to display a sentiment score for a single sentence
* as well as the individual scores of scored descriptors.
* @param sentence is the target sentence to be scored. See WordSenseBody.jsp for usage.
* @returns a String that is the sentence scored and displayed in HTML
*/
public String sumoSentimentDisplay(String sentence) {
try {
StringBuffer result = new StringBuffer();
Integer overallSentiment = DB.computeSentiment(sentence); //returns a single Integer of total sentiment
String newSentence = WordNet.wn.removeStopWords(sentence.trim());
newSentence = StringUtil.removePunctuation(newSentence);
ArrayList words = splitToArrayList(newSentence);
HashMap scoreMap = new HashMap();
for (String w : words) { //must gather individual scores that went into created the overall sentiment
Integer score = DB.computeSentimentForWord(w);
if (score != 0)
scoreMap.put(w,score);
}
result.append("Overall Sentence Sentiment Score: " + overallSentiment.toString() + "
");
if (scoreMap.size()!=0)
result.append("Individual (un-disambiguated) Word Sentiment Scores: " + scoreMap.toString() + "
");
else
result.append("All Words Are Neutral
");
return result.toString();
}
catch (NullPointerException E) {
return " NullPointerException ";
}
}
/** ***************************************************************
* A routine which takes a full pathname as input and returns a sentence by sentence display of sense
* and sentiment analysis
* @param pathname
* @param counter is used to keep track of which sentence is being displayed
* @param params is the set of html parameters
* @returns a String that is the file split into sentences, which are taken apart one by one
* and displayed in HTML
*/
public String sumoFileDisplay(String pathname, String counter, String params) {
try {
FileInputStream fileStream = new FileInputStream(pathname);
DataInputStream dataStream = new DataInputStream(fileStream);
BufferedReader reader = new BufferedReader(new InputStreamReader(dataStream));
String line;
StringBuffer textBuffer = new StringBuffer();
while ((line = reader.readLine()) != null) {
textBuffer.append(line + " ");
}
dataStream.close();
String text = textBuffer.toString();
Integer overallSentiment = DB.computeSentiment(text); //returns an Integer that is the overall sentiment of the entire file
ArrayList sentenceList = (ArrayList) splitToArrayListSentence(text); //splits by periods. Runs into trouble with titles such as Mr. Test.
StringBuffer result = new StringBuffer();
result.append("Overall File Sentiment Score: " + overallSentiment.toString() + "
");
if (Integer.parseInt(counter) > 0) {
String prevCount = Integer.toString(Integer.parseInt(counter) - 1);
result.append("" + "previous sentence (" + counter + " more)
");
}
String targetSentence = sentenceList.get(Integer.parseInt(counter));
result.append(sumoSentenceDisplay(targetSentence, text, params)); //sumoSentenceDisplay does all the heavy lifting in sentence parsing and sense gathering
if (sentenceList.size() > (Integer.parseInt(counter)+1)) {
String nextCount = Integer.toString(Integer.parseInt(counter) + 1);
String remainingCount = Integer.toString(sentenceList.size() - Integer.parseInt(counter) - 1);
result.append("" + "next sentence (" + remainingCount + " more)");
}
return result.toString();
}
catch (FileNotFoundException e) {
return " No such File ";
}
catch (IOException i) {
return " IO error ";
}
}
/** ***************************************************************
* @return true if the input String is a file pathname. Determined by whether
* the string contains a forward or backward slash. This is only used in WordSense.jsp and will fail if
* a sentence that is not a file contains a forward or back slash.
*/
public boolean isFile(String s) {
return (s.contains("\\")||s.contains("/") ? true : false);
}
/** ***************************************************************
* @return true if the first POS-prefixed synset is a hyponym of the
* second POS-prefixed synset.
* This is a recursive method.
*/
public boolean isHyponymRecurse(String synset, String hypo, ArrayList visited) {
//System.out.println("INFO in WordNet.isHyponym(): synset, hypo: " + synset + "," + hypo);
//System.out.println("INFO in WordNet.isHyponym(): synset, hypo: " + synsetsToWords.get(synset) + "," + synsetsToWords.get(hypo));
// hypernym 100021265 - food, nutrient
// public Hashtable> relations = new Hashtable();
// ~ is hyponym
if (StringUtil.emptyString(synset) || StringUtil.emptyString(hypo))
return false;
if (visited.contains(synset)) // catch cycles
return false;
ArrayList links = relations.get(synset);
if (links != null) {
for (int i = 0; i < links.size(); i++) {
AVPair link = links.get(i);
if (link == null)
System.out.println("Error in WordNet.isHyponym(): null link");
else if (link.attribute.equals("hypernym")) {
if (link.value.equals(hypo))
return true;
else {
ArrayList newVisited = new ArrayList();
newVisited.addAll(visited);
newVisited.add(synset);
if (isHyponymRecurse(link.value,hypo,newVisited))
return true;
}
}
}
}
return false;
}
/** ***************************************************************
* @return true if the first POS-prefixed synset is a hyponym of the
* second POS-prefixed synset.
* This is a recursive method.
*/
public boolean isHyponym(String synset, String hypo) {
ArrayList visited = new ArrayList();
return isHyponymRecurse(synset,hypo,visited);
}
/** ***************************************************************
* Remove stop words from a sentence.
*/
public String removeStopWords(String sentence) {
if (StringUtil.emptyString(sentence))
return "";
String result = "";
ArrayList al = splitToArrayList(sentence);
if (al == null)
return "";
for (int i = 0; i < al.size(); i++) {
String word = (String) al.get(i);
if (!stopwords.contains(word.toLowerCase())) {
if (result == "")
result = word;
else
result = result + " " + word;
}
}
return result;
}
/** ***************************************************************
* Check whether the word is a stop word
*/
public boolean isStopWord(String word) {
if (StringUtil.emptyString(word))
return false;
if (stopwords.contains(word.trim().toLowerCase()))
return true;
return false;
}
/** ***************************************************************
public String matchMultiWord(ArrayList al) {
return "";
}
*/
/** ***************************************************************
* Collect all the synsets that represent the best guess at
* meanings for all the words in a sentence. Keep track of how many
* times each sense appears.
*/
public HashMap collectCountedWordSenses(String sentence) {
if (StringUtil.emptyString(sentence))
System.out.println("Error in collectCountedWordSenses(): empty string");
HashMap result = new HashMap();
//System.out.println("INFO in collectSUMOWordSenses(): unprocessed sentence: " + sentence);
String newSentence = StringUtil.removeHTML(sentence);
newSentence = StringUtil.removePunctuation(sentence);
newSentence = removeStopWords(newSentence);
//System.out.println("INFO in collectSUMOWordSenses(): processed sentence: " + newSentence);
ArrayList al = splitToArrayList(newSentence);
if (al == null)
return result;
for (int i = 0; i < al.size(); i++) {
String word = al.get(i);
String synset = WSD.findWordSenseInContext(word,al);
if (synset != null && synset != "") {
if (result.get(synset) == null)
result.put(synset,new Integer(1));
else
result.put(synset,new Integer(result.get(synset).intValue() + 1));
}
else {
synset = WSD.getBestDefaultSense(word);
if (!StringUtil.emptyString(synset)) {
if (result.get(synset) == null)
result.put(synset,new Integer(1));
else
result.put(synset,new Integer(result.get(synset).intValue() + 1));
}
}
/**if (SUMO == null || SUMO == "")
System.out.println("INFO in findSUMOWordSense(): word not found: " + word);
else
System.out.println("INFO in findSUMOWordSense(): word, term: " + word + ", " + SUMO);*/
}
return result;
}
/** ***************************************************************
* Read the WordNet files only on initialization of the class.
*/
public static void initOnce() {
try {
if (initNeeded == true) {
if ((WordNet.baseDir == "") || (WordNet.baseDir == null))
WordNet.baseDir = KBmanager.getMgr().getPref("kbDir") + File.separator + "WordNetMappings";
baseDirFile = new File(WordNet.baseDir);
wn = new WordNet();
wn.compileRegexPatterns();
wn.readNouns();
wn.readVerbs();
wn.readAdjectives();
wn.readAdverbs();
wn.readWordCoFrequencies();
wn.readStopWords();
wn.readSenseIndex();
wn.readSenseCount();
initNeeded = false;
DB.readSentimentArray();
}
}
catch (Exception ex) {
System.out.println(ex.getMessage());
ex.printStackTrace();
}
}
/** ***************************************************************
* Split apart the block of synsets, and return the separated values
* as an array.
*/
private static String[] splitSynsets(String synsetBlock) {
String[] synsetList = null;
if (synsetBlock != null)
synsetList = synsetBlock.split("\\s+");
return synsetList;
}
/** ***************************************************************
* The main routine which looks up the search word in the hashtables
* to find the relevant word definitions and SUMO mappings.
* @param word is the word the user is asking to search for.
* @param type is whether the word is a noun or verb (we need to add capability for adjectives and adverbs.
* @param params is the set of html parameters
*/
private String sumoDisplay(String synsetBlock, String word, String type,
String sumokbname, String synsetNum, String params) {
StringBuffer result = new StringBuffer();
String synset;
String documentation = new String();
String sumoEquivalent = new String();
String kbString = "&kb=" + sumokbname;
int listLength;
String[] synsetList = splitSynsets(synsetBlock);
if (synsetList != null)
listLength = synsetList.length;
else
listLength = 0;
result.append("According to WordNet, the " + type + " \"" + word + "\" has ");
result.append(String.valueOf(listLength) + " sense(s).\n\n");
for (int i=0; i");
if (type.compareTo("noun") == 0) {
documentation = (String) nounDocumentationHash.get(synset);
result.append("1" + synset + " ");
result.append(" " + documentation + ".\n");
sumoEquivalent = (String) nounSUMOHash.get(synset);
}
else {
if (type.compareTo("verb") == 0) {
documentation = (String) verbDocumentationHash.get(synset);
result.append("2" + synset + " ");
result.append(" " + documentation + ".\n");
sumoEquivalent = (String) verbSUMOHash.get(synset);
}
else {
if (type.compareTo("adjective") == 0) {
documentation = (String) adjectiveDocumentationHash.get(synset);
result.append("3" + synset + " ");
result.append(" " + documentation + ".\n");
sumoEquivalent = (String) adjectiveSUMOHash.get(synset);
}
else {
if (type.compareTo("adverb") == 0) {
documentation = (String) adverbDocumentationHash.get(synset);
result.append("4" + synset + " ");
result.append(" " + documentation + ".\n");
sumoEquivalent = (String) adverbSUMOHash.get(synset);
}
}
}
}
if (synset.equals(synsetNum)) {
result.append("");
}
if (sumoEquivalent == null) {
result.append("- " + word + " not yet mapped to SUMO
");
}
else {
result.append(HTMLformatter.termMappingsList(sumoEquivalent,"Explore the word "+ word + " on the WordNet web site.
\n");
return result.toString();
}
/** ***************************************************************
* Return the root form of the noun, or null if it's not in the lexicon.
*/
public String nounRootForm(String mixedCase, String input) {
String result = null;
//System.out.println("INFO in WordNet.nounRootForm: Checking word : " + mixedCase + " and " + input);
if ((exceptionNounHash.containsKey(mixedCase)) ||
(exceptionNounHash.containsKey(input))) {
if (exceptionNounHash.containsKey(mixedCase))
result = (String) exceptionNounHash.get(mixedCase);
else
result = (String) exceptionNounHash.get(input);
}
else {
// Test all regular plural forms, and correct to singular.
if (WordNetUtilities.substTest(input,"s$","",nounSynsetHash))
result = WordNetUtilities.subst(input,"s$","");
else {
if (WordNetUtilities.substTest(input,"ses$","s", nounSynsetHash))
result = WordNetUtilities.subst(input,"ses$","s");
else {
if (WordNetUtilities.substTest(input,"xes$","x", nounSynsetHash))
result = WordNetUtilities.subst(input,"xes$","x");
else {
if (WordNetUtilities.substTest(input,"zes$","z", nounSynsetHash))
result = WordNetUtilities.subst(input,"zes$","z");
else {
if (WordNetUtilities.substTest(input,"ches$","ch", nounSynsetHash))
result = WordNetUtilities.subst(input,"ches$","ch");
else {
if (WordNetUtilities.substTest(input,"shes$","sh", nounSynsetHash))
result = WordNetUtilities.subst(input,"shes$","sh");
else {
if (nounSynsetHash.containsKey(mixedCase))
result = mixedCase;
else {
if (nounSynsetHash.containsKey(input))
result = input;
}
}
}
}
}
}
}
}
return result;
}
/** ***************************************************************
* This routine converts a noun to its singular form and gets the synsets for it,
* then passes those synsets to sumoDisplay() for processing.
* First check to see if the input value or its lower-case version are entered in the
* WordNet exception list (NOUN.EXC). If so, then use the regular form in the exception
* list to find the synsets in the NOUN.DAT file.
* If the word is not in the exception list, check to see if the lower case version of
* the input value is a plural and search over NOUN.DAT in the singular form if it is.
* Note that multi-word synsets must have underscores in place of spaces.
*/
private String processNoun (String sumokbname, String mixedCase, String input,
String synset, String params) {
String regular = null;
String synsetBlock;
regular = nounRootForm(mixedCase,input);
//System.out.println("Info in WordNet.processNoun(): root form: " + regular);
if (regular != null) {
synsetBlock = nounSynsetHash.get(regular);
return sumoDisplay(synsetBlock,mixedCase,"noun",sumokbname,synset,params);
}
else
return "
There are no associated SUMO terms for the noun \"" + mixedCase + "\".
\n";
}
/** ***************************************************************
* Return the present tense singular form of the verb, or null if
* it's not in the lexicon.
*/
public String verbRootForm(String mixedCase, String input) {
String result = null;
//System.out.println("INFO in verbRootForm(): word: " + mixedCase + " " + input);
if ((exceptionVerbHash.containsKey(mixedCase)) ||
(exceptionVerbHash.containsKey(input))) {
//System.out.println("INFO in verbRootForm(): exception ");
if (exceptionVerbHash.containsKey(mixedCase))
result = (String) exceptionVerbHash.get(mixedCase);
else
result = (String) exceptionVerbHash.get(input);
}
else {
//System.out.println("INFO in verbRootForm(): not an exception: " + mixedCase);
// Test all regular forms and convert to present tense singular.
if (WordNetUtilities.substTest(input,"s$","",verbSynsetHash)) {
result = WordNetUtilities.subst(input,"s$","");
//System.out.println("INFO in verbRootForm(): word: " + result);
}
else {
if (WordNetUtilities.substTest(input,"es$","",verbSynsetHash))
result = WordNetUtilities.subst(input,"es$","");
else {
if (WordNetUtilities.substTest(input,"ies$","y",verbSynsetHash))
result = WordNetUtilities.subst(input,"ies$","y");
//else if (WordNetUtilities.substTest(input,"ied$","y",verbSynsetHash)) // new 10/29/2014
// result = WordNetUtilities.subst(input,"ied$","y");
else if (WordNetUtilities.substTest(input,"ed$","",verbSynsetHash))
result = WordNetUtilities.subst(input,"ed$","");
else if (WordNetUtilities.substTest(input,"ed$","e",verbSynsetHash))
result = WordNetUtilities.subst(input,"ed$","e");
else if (WordNetUtilities.substTest(input,"ing$","e",verbSynsetHash))
result = WordNetUtilities.subst(input,"ing$","e");
else if (WordNetUtilities.substTest(input,"ing$","",verbSynsetHash))
result = WordNetUtilities.subst(input,"ing$","");
else if (verbSynsetHash.containsKey(mixedCase))
result = mixedCase;
else if (verbSynsetHash.containsKey(input))
result = input;
}
}
}
//System.out.println("INFO in verbRootForm(): result: " + result);
return result;
}
/** ***************************************************************
* This routine converts a verb to its present tense singular form and gets the synsets for it,
* then passes those synsets to sumoDisplay() for processing.
* First check to see if the input value or its lower-case version are entered in the
* WordNet exception list (VERB.EXC). If so, then use the regular form in the exception
* list to find the synsets in the VERB.DAT file.
* If the word is not in the exception list, check to see if the lower case version of the
* input value is a singular form and search over VERB.DAT with the infinitive form if it is.
* Note that multi-word synsets must have underscores in place of spaces.
*/
private String processVerb(String sumokbname, String mixedCase, String input,
String synset, String params) {
String regular = null;
String synsetBlock;
regular = verbRootForm(mixedCase,input);
System.out.println("INFO in processVerb(): word: " + regular);
if (regular != null) {
synsetBlock = verbSynsetHash.get(regular);
return sumoDisplay(synsetBlock, mixedCase, "verb", sumokbname,synset,params);
}
else
return "
There are no associated SUMO terms for the verb \"" + mixedCase + "\".
\n";
}
/** ***************************************************************
* This routine gets the synsets for an adverb, then passes those
* synsets to sumoDisplay() for processing.
* Note that multi-word synsets must have underscores in place of spaces.
*/
private String processAdverb(String sumokbname, String mixedCase, String input,
String synset, String params) {
StringBuffer result = new StringBuffer();
String synsetBlock;
synsetBlock = adverbSynsetHash.get(input);
result.append(sumoDisplay(synsetBlock, mixedCase, "adverb", sumokbname,synset,params));
return (result.toString());
}
/** ***************************************************************
* This routine gets the synsets for an adjective, then passes those
* synsets to sumoDisplay() for processing.
* Note that multi-word synsets must have underscores in place of spaces.
*/
private String processAdjective(String sumokbname, String mixedCase,
String input, String synset, String params) {
StringBuffer result = new StringBuffer();
String synsetBlock;
synsetBlock = adjectiveSynsetHash.get(input);
result.append(sumoDisplay(synsetBlock, mixedCase, "adjective", sumokbname,synset,params));
return (result.toString());
}
/** ***************************************************************
* Get all the synsets for a given word.
* @return a TreeMap of word keys and values that are ArrayLists of
* synset Strings
*/
public TreeMap> getSensesFromWord(String word) {
TreeMap> result = new TreeMap>();
String verbRoot = verbRootForm(word,word.toLowerCase());
String nounRoot = nounRootForm(word,word.toLowerCase());
ArrayList senses = wordsToSenses.get(verbRoot);
if (senses != null) {
for (int i = 0; i < senses.size(); i++) {
String sense = (String) senses.get(i); // returns a word_POS_num
if (sense != null) {
String POS = WordNetUtilities.getPOSfromKey(sense);
if (POS != null) {
String synset = WordNetUtilities.posLettersToNumber(POS) + ((String) senseIndex.get(sense));
if (synset != null) {
ArrayList words = synsetsToWords.get(synset);
if (words != null) {
Iterator it2 = words.iterator();
while (it2.hasNext()) {
String newword = (String) it2.next();
ArrayList al = result.get(newword);
if (al == null) {
al = new ArrayList();
result.put(newword,al);
}
al.add(synset);
}
}
}
}
}
}
}
senses = wordsToSenses.get(nounRoot);
if (senses != null) {
for (int i = 0; i < senses.size(); i++) {
String sense = (String) senses.get(i); // returns a word_POS_num
if (sense != null) {
String POS = WordNetUtilities.getPOSfromKey(sense);
if (POS != null) {
String synset = WordNetUtilities.posLettersToNumber(POS) + ((String) senseIndex.get(sense));
if (synset != null) {
ArrayList words = synsetsToWords.get(synset);
if (words != null) {
Iterator it2 = words.iterator();
while (it2.hasNext()) {
String newword = (String) it2.next();
ArrayList al = result.get(newword);
if (al == null) {
al = new ArrayList();
result.put(newword,al);
}
al.add(synset);
}
}
}
}
}
}
}
return result;
}
/** ***************************************************************
* Get the words and synsets corresponding to a SUMO term. The
* return is a Map of words with their corresponding synset number.
*/
public TreeMap getWordsFromTerm(String SUMOterm) {
TreeMap result = new TreeMap();
ArrayList synsets = SUMOHash.get(SUMOterm);
if (synsets == null) {
System.out.println("INFO in WordNet.getWordsFromTerm(): No synsets for term : " + SUMOterm);
return null;
}
Iterator it = synsets.iterator();
while (it.hasNext()) {
String synset = (String) it.next();
ArrayList words = synsetsToWords.get(synset);
if (words == null) {
System.out.println("INFO in WordNet.getWordsFromTerm(): No words for synset: " + synset);
return null;
}
Iterator it2 = words.iterator();
while (it2.hasNext()) {
String word = (String) it2.next();
result.put(word,synset);
}
}
return result;
}
/** ***************************************************************
*/
public ArrayList getWordsFromSynset(String synset) {
return WordNet.wn.synsetsToWords.get(synset);
}
/** ***************************************************************
* Get the SUMO term for the given root form word and part of speech.
*/
public String getSUMOterm(String word, int pos) {
if (StringUtil.emptyString(word))
return null;
String synsetBlock = null; // A String of synsets, which are 8 digit numbers, separated by spaces.
//System.out.println("INFO in WordNet.getSUMOterm: Checking word : " + word);
if (pos == NOUN)
synsetBlock = (String) nounSynsetHash.get(word);
if (pos == VERB)
synsetBlock = (String) verbSynsetHash.get(word);
if (pos == ADJECTIVE)
synsetBlock = (String) adjectiveSynsetHash.get(word);
if (pos == ADVERB)
synsetBlock = (String) adverbSynsetHash.get(word);
//int listLength;
String synset;
String[] synsetList = null;
if (synsetBlock != null)
synsetList = synsetBlock.split("\\s+");
String term = null;
if (synsetList != null) {
synset = synsetList[0]; // Just get the first synset. This needs to be changed to a word sense disambiguation algorithm.
synset = synset.trim();
if (pos == NOUN)
term = (String) nounSUMOHash.get(synset);
if (pos == VERB)
term = (String) verbSUMOHash.get(synset);
if (pos == ADJECTIVE)
term = (String) adjectiveSUMOHash.get(synset);
if (pos == ADVERB)
term = (String) adverbSUMOHash.get(synset);
}
if (term != null)
return term.trim().substring(2,term.trim().length() - 1);
else
return null;
}
/** ***************************************************************
* Does WordNet contain the given word.
*/
public boolean containsWord(String word, int pos) {
System.out.println("INFO in WordNet.containsWord: Checking word : " + word);
if (pos == NOUN && nounSynsetHash.containsKey(word))
return true;
if (pos == VERB && verbSynsetHash.containsKey(word))
return true;
if (pos == ADJECTIVE && adjectiveSynsetHash.containsKey(word))
return true;
if (pos == ADVERB && adverbSynsetHash.containsKey(word))
return true;
else
return false;
}
/** ***************************************************************
* This is the regular point of entry for this class. It takes the
* word the user is searching for, and the part of speech index, does
* the search, and returns the string with HTML formatting codes to
* present to the user. The part of speech codes must be the same as in
* the menu options in WordNet.jsp and Browse.jsp
*
* @param inp The string the user is searching for.
* @param pos The part of speech of the word 1=noun, 2=verb, 3=adjective, 4=adverb
* @return A string contained the HTML formatted search result.
*/
public String page (String inp, int pos, String kbname, String synset, String params) {
String input = inp;
StringBuffer buf = new StringBuffer();
String mixedCase = input;
String[] s = input.split("\\s+");
StringBuffer sb = new StringBuffer();
for (int i = 0; i < s.length; i++) {
sb.append(s[i]);
if ((i+1) < s.length) {
sb.append("_");
}
}
mixedCase = sb.toString();
input = sb.toString().toLowerCase();
if (pos == NOUN)
buf.append(processNoun(kbname,mixedCase,input,synset,params));
if (pos == VERB)
buf.append(processVerb(kbname,mixedCase,input,synset,params));
if (pos == ADJECTIVE)
buf.append(processAdjective(kbname,mixedCase,input,synset,params));
if (pos == ADVERB)
buf.append(processAdverb(kbname,mixedCase,input,synset,params));
buf.append("\n");
return buf.toString();
}
/** ***************************************************************
* @param synset is a synset with POS-prefix
*/
public String getDocumentation (String synset) {
char POS = synset.charAt(0);
String bareSynset = synset.substring(1);
switch (POS) {
case '1': return (String) nounDocumentationHash.get(bareSynset);
case '2': return (String) verbDocumentationHash.get(bareSynset);
case '3': return (String) adjectiveDocumentationHash.get(bareSynset);
case '4': return (String) adverbDocumentationHash.get(bareSynset);
}
return null;
}
/** ***************************************************************
* @param synset is a synset with POS-prefix
*/
public String displaySynset (String sumokbname, String synset, String params) {
StringBuffer buf = new StringBuffer();
char POS = synset.charAt(0);
String gloss = "";
String SUMOterm = "";
String POSstring = "";
String bareSynset = synset.substring(1);
switch (POS) {
case '1': gloss = (String) nounDocumentationHash.get(bareSynset);
SUMOterm = (String) nounSUMOHash.get(bareSynset);
POSstring = "Noun";
break;
case '2': gloss = (String) verbDocumentationHash.get(bareSynset);
SUMOterm = (String) verbSUMOHash.get(bareSynset);
POSstring = "Verb";
break;
case '3': gloss = (String) adjectiveDocumentationHash.get(bareSynset);
SUMOterm = (String) adjectiveSUMOHash.get(bareSynset);
POSstring = "Adjective";
break;
case '4': gloss = (String) adverbDocumentationHash.get(bareSynset);
SUMOterm = (String) adverbSUMOHash.get(bareSynset);
POSstring = "Adverb";
break;
}
if (gloss == null) {
return (synset + " is not a valid synset number.\n");
}
buf.append("" + POSstring + " Synset: " + synset);
if (SUMOterm != null && SUMOterm != "")
buf.append(HTMLformatter.termMappingsList(SUMOterm,"Words: ");
Iterator it = words.iterator();
while (it.hasNext()) {
String word = it.next();
buf.append(word);
if (it.hasNext())
buf.append(", ");
}
buf.append("\n Gloss: " + gloss);
buf.append("
\n");
ArrayList al3 = relations.get(synset);
if (al3 != null) {
Iterator it3 = al3.iterator();
while (it3.hasNext()) {
AVPair avp = it3.next();
buf.append(avp.attribute + " ");
buf.append("" + avp.value + " - ");
words = new TreeSet();
ArrayList al2 = synsetsToWords.get(avp.value);
if (al2 != null)
words.addAll(al2);
Iterator it2 = words.iterator();
while (it2.hasNext()) {
String word = it2.next();
buf.append(word);
if (it2.hasNext())
buf.append(", ");
}
buf.append("
\n");
}
buf.append("\n");
}
return buf.toString();
}
/** ***************************************************************
* @param key is a WordNet sense key
* @return 9-digit POS-prefix and synset number
*/
public String displayByKey (String sumokbname, String key, String params) {
String synset = (String) senseIndex.get(key);
String POS = WordNetUtilities.getPOSfromKey(key);
POS = POS.substring(0,1);
return displaySynset(sumokbname,POS+synset,params);
}
/** ***************************************************************
private String fromXML(SimpleElement configuration) {
StringBuffer result = new StringBuffer();
if (!configuration.getTagName().equals("wordnet"))
System.out.println("Error in KBmanager.fromXML(): Bad tag: " + configuration.getTagName());
else {
}
return result.toString();
}
*/
/** ***************************************************************
*/
private SimpleElement toXML() {
SimpleElement top = new SimpleElement("wordnet");
Iterator it = synsetsToWords.keySet().iterator();
while (it.hasNext()) {
String synset = (String) it.next();
String gloss = "";
String name = "";
String SUMO = "";
switch (synset.charAt(0)) {
case '1': gloss = (String) nounDocumentationHash.get(synset.substring(1));
SUMO = (String) nounSUMOHash.get(synset.substring(1));
break;
case '2': gloss = (String) verbDocumentationHash.get(synset.substring(1));
SUMO = (String) verbSUMOHash.get(synset.substring(1));
break;
case '3': gloss = (String) adjectiveDocumentationHash.get(synset.substring(1));
SUMO = (String) adjectiveSUMOHash.get(synset.substring(1));
break;
case '4': gloss = (String) adverbDocumentationHash.get(synset.substring(1));
SUMO = (String) adverbSUMOHash.get(synset.substring(1));
break;
}
if (gloss != null)
gloss = gloss.replaceAll("\"",""e;");
ArrayList al2 = synsetsToWords.get(synset);
Iterator itts = al2.iterator();
if (itts.hasNext())
name = (String) itts.next();
SimpleElement item = new SimpleElement("item");
item.setAttribute("id",synset);
item.setAttribute("offset",synset);
item.setAttribute("type","synset");
item.setAttribute("name",name);
item.setAttribute("source","WordNet 2.0");
item.setAttribute("gloss",gloss);
top.addChildElement(item);
ArrayList al = relations.get(synset);
if (SUMO != "") {
String bareTerm = WordNetUtilities.getBareSUMOTerm(SUMO);
char mapping = WordNetUtilities.getSUMOMappingSuffix(SUMO);
SimpleElement link = new SimpleElement("link");
link.setAttribute("id1",synset);
link.setAttribute("id2",bareTerm);
String mapName = WordNetUtilities.mappingCharToName(mapping);
link.setAttribute("type",mapName);
top.addChildElement(link);
}
if (al != null) {
Iterator it2 = al.iterator();
while (it2.hasNext()) {
SimpleElement link = new SimpleElement("link");
AVPair avp = it2.next();
link.setAttribute("type",avp.attribute);
link.setAttribute("id1",synset);
link.setAttribute("id2",avp.value);
top.addChildElement(link);
}
}
}
return top;
}
/** ***************************************************************
*/
public void writeXML() {
FileWriter fw = null;
PrintWriter pw = null;
String dir = WordNet.baseDir;
String fname = "WordNet.xml";
try {
fw = new FileWriter(dir + File.separator + fname);
pw = new PrintWriter(fw);
SimpleElement se = toXML();
pw.println(se.toFileString());
}
catch (Exception e) {
System.out.println("Error writing file " + dir + File.separator + fname + "\n" + e.getMessage());
e.printStackTrace();
}
finally {
try {
if (pw != null) {
pw.close();
}
if (fw != null) {
fw.close();
}
}
catch (Exception ex) {
}
}
}
/** *************************************************************
*/
private static boolean arrayContains(int[] ar, int value) {
//System.out.println("INFO in WordNet.arrayContains: value: " + value);
for (int i = 0; i < ar.length; i++) {
if (ar[i] == value)
return true;
}
return false;
}
/** *************************************************************
* Frame transitivity
* intransitive - 1,2,3,4,7,23,35
* transitive - everything else
* ditransitive - 15,16,17,18,19
*/
private String getTransitivity(String synset, String word) {
//System.out.println("INFO in WordNet.getTransitivity: synset, word: " + synset + " " + word);
int[] intrans = {1,2,3,4,7,23,35};
int[] ditrans = {15,16,17,18,19};
String intransitive = "no";
String transitive = "no";
String ditransitive = "no";
ArrayList frames = new ArrayList();
ArrayList res = verbFrames.get(synset);
if (res != null)
frames.addAll(res);
res = verbFrames.get(synset + "-" + word);
if (res != null)
frames.addAll(res);
for (int i = 0; i < frames.size(); i++) {
int value = (Integer.valueOf((String) frames.get(i))).intValue();
if (arrayContains(intrans,value))
intransitive = "intransitive";
else
if (arrayContains(ditrans,value))
ditransitive = "ditransitive";
else
transitive = "transitive";
}
return "[" + intransitive + "," + transitive + "," + ditransitive + "]";
}
/** *************************************************************
* Replace underscores with commas, wrap hyphenatid and apostrophed words in single
* quotes, and wrap the whole phrase in brackets. Used only in conversion
* to Prolog.
*/
private static String processMultiWord(String word) {
word = word.replace('_',',');
word = word.replace("'","\\'");
String words[] = word.split(",");
word = "";
for (int i = 0; i < words.length; i++) {
if (words[i].length() > 0 &&
(words[i].indexOf("-") > -1 || (words[i].indexOf(".") > -1) ||
(words[i].indexOf("\\'") > -1) || Character.isUpperCase(words[i].charAt(0)) || Character.isDigit(words[i].charAt(0))))
words[i] = "'" + words[i] + "'";
word = word + words[i];
if (i < words.length-1)
word = word + ",";
}
return "[" + word + "]";
}
/** *************************************************************
*/
private static String multipleMappingToProlog(String term) {
if (term.indexOf("&%") < 0)
return term;
else {
term = "[" + term + "]";
term = term.replaceAll("[\\=\\+] \\&\\%","','");
}
return term;
}
/** *************************************************************
* verb_in_lexicon(Verb for singular mode, Verb for plural mode, {transitive,
* intransitive, [intransitive, transitive, ditransitive], [no, no,
* ditransitive], [no, transitive, no], [intransitive, no, no], [no,
* transitive, ditransitive], [intransitive, transitive, no], [no, no, no],
* [intransitive, no, ditransitive]}, singular, {simple, prepositional,
* compound, phrasal}, {event, state}, SUMOMapping., Synset_ID).
*/
private void writeVerbsProlog(PrintWriter pw, KB kb) throws IOException {
Iterator it = verbSynsetHash.keySet().iterator();
while (it.hasNext()) {
String word = (String) it.next();
String compound = "simple";
if (word.indexOf("_") > -1)
compound = "compound";
String stringSynsets = (String) verbSynsetHash.get(word);
String plural = WordNetUtilities.verbPlural(word);
if (word.indexOf("_") > -1) {
word = processMultiWord(word);
plural = processMultiWord(plural);
}
else {
word = word.replace("'","\\'");
if (word.indexOf("-") > -1 || (word.indexOf(".") > -1) ||
(word.indexOf("\\'") > -1) || Character.isUpperCase(word.charAt(0)) || Character.isDigit(word.charAt(0))) {
word = "'" + word + "'";
plural = "'" + plural + "'";
}
}
String[] synsetList = splitSynsets(stringSynsets);
//Iterator it2 = verbSUMOHash.keySet().iterator();
for (int i = 0; i < synsetList.length; i++) {
String synset = synsetList[i];
String sumoTerm = (String) verbSUMOHash.get(synset);
if (sumoTerm != null && sumoTerm != "") {
String bareSumoTerm = WordNetUtilities.getBareSUMOTerm(sumoTerm);
String transitivity = getTransitivity(synset,word);
String eventstate = "state";
if (kb.childOf(bareSumoTerm,"Process"))
eventstate = "event";
bareSumoTerm = multipleMappingToProlog(bareSumoTerm);
pw.println("verb_in_lexicon(" + plural + "," + word + "," + transitivity +
", singular, " + compound + ", " + eventstate + ", '" + bareSumoTerm + "',2" +
synset + ").");
}
}
}
}
/** *************************************************************
* adjective_in_lexicon(Adj, CELT_form, {normal, two_place}, {positive,
* ungraded, comparative, superlative}, SUMOMapping).
*/
private void writeAdjectivesProlog(PrintWriter pw, KB kb) throws IOException {
Iterator it = adjectiveSynsetHash.keySet().iterator();
while (it.hasNext()) {
String word = (String) it.next();
//String compound = "simple";
//if (word.indexOf("_") > -1)
// compound = "compound";
String stringSynsets = (String) adjectiveSynsetHash.get(word);
if (word.indexOf("_") > -1) {
word = processMultiWord(word);
}
else {
word = word.replace("'","\\'");
if (word.indexOf("-") > -1 || (word.indexOf(".") > -1) ||
(word.indexOf("\\'") > -1) || Character.isUpperCase(word.charAt(0)) || Character.isDigit(word.charAt(0))) {
word = "'" + word + "'";
}
}
String[] synsetList = splitSynsets(stringSynsets);
//Iterator it2 = adjectiveSUMOHash.keySet().iterator();
for (int i = 0; i < synsetList.length; i++) {
String synset = synsetList[i];
String sumoTerm = (String) adjectiveSUMOHash.get(synset);
if (sumoTerm != null && sumoTerm != "") {
String bareSumoTerm = WordNetUtilities.getBareSUMOTerm(sumoTerm);
bareSumoTerm = multipleMappingToProlog(bareSumoTerm);
pw.println("adjective_in_lexicon(" + word + "," + word + ",normal,positive," +
bareSumoTerm + ").");
}
}
}
}
/** *************************************************************
* adverb_in_lexicon(Adv, {location, direction, time, duration, frequency,
* manner}, SUMOMapping).
*/
private void writeAdverbsProlog(PrintWriter pw, KB kb) throws IOException {
Iterator it = verbSynsetHash.keySet().iterator();
while (it.hasNext()) {
String word = (String) it.next();
//String compound = "simple";
//if (word.indexOf("_") > -1)
// compound = "compound";
String stringSynsets = (String) verbSynsetHash.get(word);
if (word.indexOf("_") > -1) {
word = processMultiWord(word);
}
else {
word = word.replace("'","\\'");
if (word.indexOf("-") > -1 || (word.indexOf(".") > -1) ||
(word.indexOf("\\'") > -1) || Character.isUpperCase(word.charAt(0)) || Character.isDigit(word.charAt(0))) {
word = "'" + word + "'";
}
}
String[] synsetList = splitSynsets(stringSynsets);
//Iterator it2 = verbSUMOHash.keySet().iterator();
for (int i = 0; i < synsetList.length; i++) {
String synset = synsetList[i];
String sumoTerm = (String) verbSUMOHash.get(synset);
if (sumoTerm != null && sumoTerm != "") {
String bareSumoTerm = WordNetUtilities.getBareSUMOTerm(sumoTerm);
bareSumoTerm = multipleMappingToProlog(bareSumoTerm);
pw.println("adverb_in_lexicon(" + word + ",null," + bareSumoTerm + ").");
}
}
}
}
/** *************************************************************
* noun_in_lexicon(Noun,{object, person, time}, neuter, {count, mass}, singular, SUMOMapping, Synset_ID).
*/
private void writeNounsProlog(PrintWriter pw, KB kb) throws IOException {
Iterator it = nounSynsetHash.keySet().iterator();
while (it.hasNext()) {
String word = (String) it.next();
String stringSynsets = (String) nounSynsetHash.get(word);
boolean uppercase = false;
if (Character.isUpperCase(word.charAt(0)))
uppercase = true;
if (word.indexOf("_") > -1)
word = processMultiWord(word);
else {
word = word.replace("'","\\'");
if ((word.indexOf("-") > -1) || (word.indexOf(".") > -1) || (word.indexOf("/") > -1)
|| (word.indexOf("\\'") > -1) || uppercase || Character.isDigit(word.charAt(0)))
word = "'" + word + "'";
}
String[] synsetList = splitSynsets(stringSynsets);
for (int i = 0; i < synsetList.length; i++) {
String synset = synsetList[i];
String sumoTerm = (String) nounSUMOHash.get(synset);
if (sumoTerm != null && sumoTerm != "") {
String bareSumoTerm = WordNetUtilities.getBareSUMOTerm(sumoTerm);
char mapping = WordNetUtilities.getSUMOMappingSuffix(sumoTerm);
String type = "object";
if (kb.childOf(bareSumoTerm,"Human") || kb.childOf(bareSumoTerm,"SocialRole"))
type = "person";
if (kb.childOf(bareSumoTerm,"TimePosition") || kb.childOf(bareSumoTerm,"Process"))
type = "time";
String countOrMass = "count";
if (kb.childOf(bareSumoTerm,"Substance"))
countOrMass = "mass";
boolean instance = false;
if (uppercase && mapping == '@')
instance = true;
if (mapping == '=') {
ArrayList al = kb.instancesOf(bareSumoTerm);
if (al.size() > 0)
instance = true;
}
if (instance && uppercase) {
ArrayList al = kb.askWithRestriction(1,bareSumoTerm,0,"instance");
String parentTerm = "";
if (al != null && al.size() > 0) {
parentTerm = al.get(0).getArgument(2);
}
else
parentTerm = bareSumoTerm;
bareSumoTerm = multipleMappingToProlog(bareSumoTerm);
parentTerm = multipleMappingToProlog(parentTerm);
pw.println("proper_noun_in_lexicon(" + word + "," + type + ", neuter, singular, '" +
parentTerm + "','" + bareSumoTerm + "',1" + synset + ").");
}
else {
bareSumoTerm = multipleMappingToProlog(bareSumoTerm);
pw.println("noun_in_lexicon(" + word + "," + type + ", neuter, " +
countOrMass + ", singular, '" + bareSumoTerm + "',1" +
synset + ").");
}
}
}
}
}
/** ***************************************************************
*/
public void writeProlog(KB kb) {
FileWriter fw = null;
PrintWriter pw = null;
String dir = WordNet.baseDir;
String fname = "WordNet.pl";
try {
fw = new FileWriter(dir + File.separator + fname);
pw = new PrintWriter(fw);
writeNounsProlog(pw,kb);
writeVerbsProlog(pw,kb);
writeAdjectivesProlog(pw,kb);
writeAdverbsProlog(pw,kb);
}
catch (Exception e) {
System.out.println("Error writing file " + dir + File.separator + fname + "\n" + e.getMessage());
e.printStackTrace();
}
finally {
try {
if (pw != null)
pw.close();
if (fw != null)
fw.close();
}
catch (Exception ex) {
}
}
}
/** ***************************************************************
*/
public String senseKeyPOS (String senseKey) {
if (StringUtil.emptyString(senseKey))
return "";
int underscore2 = senseKey.lastIndexOf("_");
if (underscore2 < 0) return "";
int underscore1 = senseKey.lastIndexOf("_",underscore2-1);
if (underscore1 < 0) return "";
return senseKey.substring(underscore1+1, underscore2);
}
/** ***************************************************************
*/
private String senseKeySenseNum (String senseKey) {
if (senseKey == null) return "";
int underscore2 = senseKey.lastIndexOf("_");
if (underscore2 < 0) return "";
int underscore1 = senseKey.lastIndexOf("_",underscore2-1);
if (underscore1 < 0) return "";
return senseKey.substring(underscore2+1,senseKey.length());
}
/** ***************************************************************
* Find the "word number" of a word and synset, which is its place
* in the list of words belonging to a given synset. Return -1 if
* not found.
*/
private int findWordNum(String POS, String synset, String word) {
ArrayList al = synsetsToWords.get(POS+synset);
if (al == null || al.size() < 1) {
System.out.println("Error in WordNet.findWordNum(): No words found for synset: " + POS + synset + " and word " + word);
return -1;
}
for (int i = 0; i < al.size(); i++) {
String storedWord = (String) al.get(i);
if (word.equalsIgnoreCase(storedWord)) {
return i+1;
}
}
System.out.println("Error in WordNet.findWordNum(): No match found for synset: " + POS + synset + " and word " + word);
System.out.println(al);
return -1;
}
/** ***************************************************************
*/
private String processWordForProlog(String word) {
String result = new String(word);
int start = 0;
while (result.indexOf("'",start) > -1) {
int i = 0;
i = result.indexOf("'",start);
//System.out.println("INFO in WordNet.processPrologString(): index: " + i + " string: " + doc);
if (i == 0)
result = "''" + result.substring(i+1);
else
result = result.substring(0,i) + "\\'" + result.substring(i+1);
start = i+2;
}
return result;
}
/** ***************************************************************
* Write WordNet data to a prolog file with a single kind of clause
* in the following format:
* s(Synset_ID, Word_No_in_the_Synset, Word, SS_Type,
* Synset_Rank_By_the_Word,Tag_Count)
*/
public void writeWordNetS() {
FileWriter fw = null;
PrintWriter pw = null;
String dir = WordNet.baseDir;
String fname = "Wn_s.pl";
try {
fw = new FileWriter(dir + File.separator + fname);
pw = new PrintWriter(fw);
if (wordsToSenses.keySet().size() < 1)
System.out.println("Error in WordNet.writeWordNetS(): No contents in sense index");
Iterator it = wordsToSenses.keySet().iterator();
while (it.hasNext()) {
String word = (String) it.next();
String processedWord = processWordForProlog(word);
ArrayList keys = wordsToSenses.get(word);
Iterator it2 = keys.iterator();
if (keys.size() < 1)
System.out.println("Error in WordNet.writeWordNetS(): No synsets for word: " + word);
while (it2.hasNext()) {
String senseKey = (String) it2.next();
//System.out.println("INFO in WordNet.writeWordNetS(): Sense key: " + senseKey);
String POS = senseKeyPOS(senseKey);
String senseNum = senseKeySenseNum(senseKey);
if (POS == "" || senseNum == "")
System.out.println("Error in WordNet.writeWordNetS(): Bad sense key: " + senseKey);
POS = WordNetUtilities.posLettersToNumber(POS);
String POSchar = Character.toString(WordNetUtilities.posNumberToLetter(POS.charAt(0)));
String synset = (String) senseIndex.get(senseKey);
int wordNum = findWordNum(POS,synset,word);
pw.println("s(" + POS + synset + "," + wordNum + ",'" + processedWord + "'," + POSchar + "," + senseNum + ",1).");
}
}
}
catch (Exception e) {
System.out.println("Error writing file " + dir + File.separator + fname + "\n" + e.getMessage());
e.printStackTrace();
}
finally {
try {
if (pw != null)
pw.close();
if (fw != null)
fw.close();
}
catch (Exception ex) {
}
}
}
/** ***************************************************************
*/
public void writeWordNetHyp() {
FileWriter fw = null;
PrintWriter pw = null;
String dir = WordNet.baseDir;
String fname = "Wn_hyp.pl";
try {
fw = new FileWriter(dir + File.separator + fname);
pw = new PrintWriter(fw);
if (relations.keySet().size() < 1)
System.out.println("Error in WordNet.writeWordNetHyp(): No contents in relations");
Iterator it = relations.keySet().iterator();
while (it.hasNext()) {
String synset = (String) it.next();
//System.out.println("INFO in WordNet.writeWordNetHyp(): synset: " + synset);
ArrayList rels = relations.get(synset);
if (rels == null || rels.size() < 1)
System.out.println("Error in WordNet.writeWordNetHyp(): No contents in rels for synset: " + synset);
if (rels != null) {
Iterator it2 = rels.iterator();
while (it2.hasNext()) {
AVPair rel = it2.next();
if (rel.attribute.equals("hypernym"))
pw.println("hyp(" + synset + "," + rel.value + ").");
}
}
}
}
catch (Exception e) {
System.out.println("Error writing file " + dir + File.separator + fname + "\n" + e.getMessage());
e.printStackTrace();
}
finally {
try {
if (pw != null)
pw.close();
if (fw != null)
fw.close();
}
catch (Exception ex) {
}
}
}
/** ***************************************************************
* Double any single quotes that appear.
*/
public String processPrologString (String doc) {
int start = 0;
while (doc.indexOf("'",start) > -1) {
int i = 0;
i = doc.indexOf("'",start);
//System.out.println("INFO in WordNet.processPrologString(): index: " + i + " string: " + doc);
if (i == 0)
doc = "''" + doc.substring(i+1);
else
doc = doc.substring(0,i) + "''" + doc.substring(i+1);
start = i+2;
}
return doc;
}
/** ***************************************************************
*/
public void writeWordNetG() {
FileWriter fw = null;
PrintWriter pw = null;
String dir = WordNet.baseDir;
String fname = "Wn_g.pl";
try {
fw = new FileWriter(dir + File.separator + fname);
pw = new PrintWriter(fw);
Iterator it = nounDocumentationHash.keySet().iterator();
while (it.hasNext()) {
String synset = (String) it.next();
String doc = (String) nounDocumentationHash.get(synset);
doc = processPrologString(doc);
pw.println("g(" + "1" + synset + ",'(" + doc + ")').");
}
it = verbDocumentationHash.keySet().iterator();
while (it.hasNext()) {
String synset = (String) it.next();
String doc = (String) verbDocumentationHash.get(synset);
doc = processPrologString(doc);
pw.println("g(" + "2" + synset + ",'(" + doc + ")').");
}
it = adjectiveDocumentationHash.keySet().iterator();
while (it.hasNext()) {
String synset = (String) it.next();
String doc = (String) adjectiveDocumentationHash.get(synset);
doc = processPrologString(doc);
pw.println("g(" + "3" + synset + ",'(" + doc + ")').");
}
it = adverbDocumentationHash.keySet().iterator();
while (it.hasNext()) {
String synset = (String) it.next();
String doc = (String) adverbDocumentationHash.get(synset);
doc = processPrologString(doc);
pw.println("g(" + "4" + synset + ",'(" + doc + ")').");
}
}
catch (Exception e) {
System.out.println("Error writing file " + dir + File.separator + fname + "\n" + e.getMessage());
e.printStackTrace();
}
finally {
try {
if (pw != null)
pw.close();
if (fw != null)
fw.close();
}
catch (Exception ex) {
}
}
}
/** ***************************************************************
*/
public void writeWordNetProlog () throws IOException {
writeWordNetS();
writeWordNetHyp();
writeWordNetG();
}
/** ***************************************************************
* Generate a new noun synset ID that doesn't have an existing hash
*/
public String generateNounSynsetID() {
//System.out.println("INFO in WordNet.generateNounSynsetID()");
//TreeSet ts = new TreeSet();
//ts.addAll(nounSUMOHash.keySet());
String l = maxNounSynsetID;
//System.out.println("INFO in WordNet.generateNounSynsetID(): last synset: " + l);
Integer intval = Integer.parseInt(l);
intval++;
maxNounSynsetID = String.valueOf(intval);
return maxNounSynsetID;
}
/** ***************************************************************
* Generate a new noun synset ID that doesn't have an existing hash
*/
public String generateVerbSynsetID() {
//System.out.println("INFO in WordNet.generateVerbSynsetID()");
//TreeSet ts = new TreeSet();
//ts.addAll(verbSUMOHash.keySet());
String l = maxVerbSynsetID;
//System.out.println("INFO in WordNet.generateVerbSynsetID(): last synset: " + l);
Integer intval = Integer.parseInt(l);
intval++;
maxVerbSynsetID = String.valueOf(intval);
return maxVerbSynsetID;
}
/** ***************************************************************
* Generate a new noun synset from a termFormat
*/
public String nounSynsetFromTermFormat(String tf, String SUMOterm, KB kb) {
String synsetID = generateNounSynsetID();
ArrayList forms = kb.askWithRestriction(0, "documentation", 1, SUMOterm);
if (forms.size() > 0) {
Formula f = forms.get(0);
String doc = f.getArgument(3);
nounDocumentationHash.put(synsetID, doc);
}
nounSUMOHash.put(synsetID, SUMOterm + "=");
//nounSynsetHash
return synsetID;
}
/** ***************************************************************
* Generate a new verb synset from a termFormat
*/
public String verbSynsetFromTermFormat(String tf, String SUMOterm, KB kb) {
String synsetID = generateVerbSynsetID();
ArrayList forms = kb.askWithRestriction(0, "documentation", 1, SUMOterm);
if (forms.size() > 0) {
Formula f = forms.get(0);
String doc = f.getArgument(3);
verbDocumentationHash.put(synsetID, doc);
}
verbSUMOHash.put(synsetID, SUMOterm + "=");
//verbSynsetHash
return synsetID;
}
/** ***************************************************************
* Generate a new synset from a termFormat
*/
public void synsetFromTermFormat(String tf, String SUMOterm, KB kb) {
String synsetID = null;
String pos = null;
if (kb.kbCache.getParentClasses(SUMOterm) != null &&
kb.kbCache.getParentClasses(SUMOterm).contains("Process")) { // like a verb
pos = "2";
synsetID = "2" + verbSynsetFromTermFormat(tf,SUMOterm,kb);
}
else { // like a noun
pos = "1";
synsetID = "1" + nounSynsetFromTermFormat(tf,SUMOterm,kb);
}
ArrayList al = SUMOHash.get(SUMOterm);
if (al == null)
al = new ArrayList();
al.add(synsetID);
SUMOHash.put(SUMOterm, al);
ArrayList words = synsetsToWords.get(synsetID);
if (words == null)
words = new ArrayList();
words.add(tf);
synsetsToWords.put(synsetID,words);
String letterPOS = WordNetUtilities.posNumberToLetters(pos);
String key = tf + "_" + letterPOS + "_1"; // FIXME: Note! This will cause a bug if there's a name clash
senseIndex.put(key,synsetID.substring(1)); // senseIndex requires un-prefixed synset #
ArrayList keys = new ArrayList();
if (wordsToSenses.containsKey(tf))
keys = wordsToSenses.get(tf);
keys.add(key);
wordsToSenses.put(tf,keys);
}
/** ***************************************************************
* Generate a new synset from a termFormat
*/
public void termFormatsToSynsets(KB kb) {
//System.out.println("INFO in WordNet.termFormatsToSynsets()");
ArrayList forms = kb.ask("arg", 0, "termFormat");
for (int i = 0; i < forms.size(); i++) {
Formula form = forms.get(i);
String SUMOterm = form.getArgument(2);
//if (SUMOterm.equals("FourStrokeEngine"))
// System.out.println("INFO in WordNet.termFormatsToSynsets(): formula: " + form);
String tf = form.getArgument(3).substring(1); // remove first quote
tf = tf.substring(0,tf.length()-1); // remove last quote
tf = tf.replaceAll(" ", "_");
if (tf.indexOf("_") > -1) {
//System.out.println("INFO in WordNet.termFormatsToSynsets() multiword:" + tf + " SUMO: " + SUMOterm);
multiWords.addMultiWord(tf);
}
synsetFromTermFormat(tf,SUMOterm,kb);
}
}
/** ***************************************************************
* A method used only for testing. It should not be called
* during normal operation.
*/
public static void testWordFreq () {
String word = "run";
//String word = "be";
try {
KBmanager.getMgr().initializeOnce();
} catch (Exception ex) {
System.out.println(ex.getMessage());
}
System.out.println("INFO in WordNet.main(): done initializing");
WordNet.initOnce();
System.out.println("Word frequencies: " + WordNet.wn.wordFrequencies.get(word));
System.out.println("Best word frequency: " + WordNet.wn.wordFrequencies.get(word).last());
System.out.println("Best word frequency value: " + WordNet.wn.wordFrequencies.get(word).last().value);
// HashMap> wordFrequencies
System.out.println("Best word frequency value sense index: " +
WordNet.wn.senseIndex.get(WordNet.wn.wordFrequencies.get(word).last().value));
System.out.println("Best default sense w/POS: " + WSD.getBestDefaultSense(word,2));
System.out.println("Best default sense w/o POS: " + WSD.getBestDefaultSense(word));
System.out.println("Best default sense w/o POS sense index: " +
WordNet.wn.reverseSenseIndex.get(WSD.getBestDefaultSense(word,2)));
}
/** ***************************************************************
* A method used only for testing. It should not be called
* during normal operation.
*/
public static void testProcessPointers() {
String line = "01522276 35 v 04 wind 6 wrap 2 roll 0 twine 3 013 @ 01850315 v 0000 + 03150232 n 0303 + 07441619 n 0303 ^ 00435688 v 0301 ^ 00435688 v 0202 + 04586421 n 0101 + 10781984 n 0101 ! 01523654 v 0101 ~ 01522878 v 0000 ~ 01523105 v 0000 ~ 01523270 v 0000 ~ 01523401 v 0000 ~ 01523986 v 0000 01 + 21 00 | arrange or or coil around; \"roll your hair around your finger\"; \"Twine the thread around the spool\"; \"She wrapped her arms around the child\" &%Motion+";
// 10: p = Pattern.compile("^([0-9]{8})([^\\|]+)\\|\\s([\\S\\s]+?)\\s(\\(?\\&\\%\\S+[\\S\\s]+)$");
WordNet.wn.compileRegexPatterns();
Matcher m = regexPatterns[10].matcher(line);
if (m.matches()) {
//verbDocumentationHash.put(m.group(1),m.group(3));
//addSUMOMapping(m.group(4),"2" + m.group(1));
WordNet.wn.processPointers("2" + m.group(1),m.group(2));
}
System.out.println("Info in WordNet.testProcessPointers(): synset: " + WordNet.wn.verbSynsetHash.get("roll"));
}
/** ***************************************************************
* A main method, used only for testing. It should not be called
* during normal operation.
*/
public static void main (String[] args) {
//testWordFreq();
testProcessPointers();
}
}