com.articulate.sigma.wordNet

Class WordNetUtilities

java.lang.Object

com.articulate.sigma.wordNet.WordNetUtilities

public class WordNetUtilities
extends java.lang.Object

Field Summary

Fields

Modifier and Type	Field and Description
static int	errorCount
static java.util.HashMap<java.lang.String,java.lang.String>	mappings POS-prefixed mappings from a new synset number to the old one.
static int	TPTPidCounter
static boolean	withThoughtEmotion
protected static java.util.ArrayList<java.lang.String>	WordNetRelations

Constructor Summary

Constructors

Constructor and Description
WordNetUtilities()

Method Summary

Modifier and Type	Method and Description
static java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>>	collapseSenses()
static void	commentSentiment(java.lang.String fileWithPath)
static java.util.ArrayList<java.lang.String>	convertTermList(java.lang.String termList) Convert a list of Terms in the format "&%term1 &%term2" to an ArrayList of bare term Strings
static java.lang.String	convertWordNetPointer(java.lang.String ptr)
static void	deduceMissingLinks() Use the WordNet hyper-/hypo-nym links to deduce a likely link for a SUMO term that has not yet been manually linked.
static void	extractMeronyms() A utility to extract meronym relations as relations between SUMO terms.
static java.util.HashSet<java.lang.String>	findLeaves(java.lang.String rel) Find all the leaf nodes for a particular relation in WordNet.
static java.util.HashSet<java.lang.String>	findLeavesInTree(java.util.HashSet<java.lang.String> rels) Find all the leaf nodes for a particular relation in WordNet.
static java.util.ArrayList<java.util.ArrayList<java.lang.String>>	findPathsToRoot(java.util.ArrayList<java.lang.String> base, java.lang.String synset) Find the complete path from a given synset.
static java.lang.String	formatWords(java.util.TreeMap<java.lang.String,java.lang.String> words, java.lang.String kbName) HTML format a TreeMap of word senses and their associated synset
static java.lang.String	formatWordsList(java.util.TreeMap<java.lang.String,java.util.ArrayList<java.lang.String>> words, java.lang.String kbName) HTML format a TreeMap of ArrayLists word senses
static void	generateHyponymSets(java.lang.String filename) Generate sets of all hyponymous words for each synset in a file
static void	generateSUMOfromWN() Generate notional SUMO terms from WordNet.
static void	generateSUMOfromWN(java.lang.String synset, java.lang.String sumo) Generate notional SUMO terms from WordNet
static void	generateSUMOfromWNsubtree(java.lang.String synset, java.lang.String sumo) Generate notional SUMO terms from WordNet
static java.util.HashSet<java.lang.String>	getAllHyponyms(java.lang.String s)
static java.util.HashSet<java.lang.String>	getAllHyponymsTransitive(java.lang.String s)
static java.lang.String	getBareSUMOTerm(java.lang.String term) Get a SUMO term minus its &% prefix and one character mapping suffix.
static java.lang.String	getKeyFromSense(java.lang.String synset) Get the word_POS_num sense key corresponding to a 9 digit synset.
static java.lang.String	getPOSfromKey(java.lang.String senseKey) Extract the POS from a word_POS_num sense key.
static java.lang.String	getSenseFromKey(java.lang.String senseKey) Extract the synset corresponding to a word_POS_num sense key.
static char	getSUMOMappingSuffix(java.lang.String term) Get a SUMO term mapping suffix.
static java.util.ArrayList<java.lang.String>	getSynsetsFromSUMO(java.lang.String sumo) get all synsets corresponding to a SUMO term
static java.util.HashSet<java.lang.String>	getSynsetsFromSUMOList(java.util.Collection<java.lang.String> sumo) get all synsets corresponding to a list of SUMO terms
static java.lang.String	getWordFromKey(java.lang.String senseKey) Extract the word from a word_POS_num sense key.
static java.util.HashSet<java.lang.String>	getWordsFromSynsetList(java.util.Collection<java.lang.String> synsets) get all words corresponding to a list of synsets
void	imageNetLinks() Import links from www.image-net.org that are linked to WordNet and links them to SUMO terms when the synset has a directly equivalent SUMO term
static boolean	isHyponymousWord(java.lang.String word, java.util.HashSet<java.lang.String> synsets)
static boolean	isValidKey(java.lang.String senseKey) Check whether a sense key format is valid
static boolean	isValidSynset8(java.lang.String synset) Check whether a synset format is valid
static boolean	isValidSynset9(java.lang.String synset) Check whether a synset format is valid
static java.lang.String	lowestCommonParent(java.lang.String s1, java.lang.String s2)
static void	main(java.lang.String[] args) A main method, used only for testing.
static java.lang.String	mappingCharToName(char mappingType)
static void	mergeUpdates() Read in a file with a nine-digit synset number followed by a space and a SUMO term.
static int	numSynsets(char pos)
static java.lang.String	posLettersToNumber(java.lang.String pos) Convert a part of speech number to the two letter format used by the WordNet sense index code.
static char	posLetterToNumber(char POS)
static char	posNumberToLetter(char POS)
static java.lang.String	posNumberToLetters(java.lang.String pos) Convert a part of speech number to the two letter format used by the WordNet sense index code.
static char	posPennToNumber(java.lang.String penn)
static java.lang.String	printStatistics()
static void	processMissingLinks(java.lang.String fileName, java.lang.String pattern, java.lang.String posNum) This is a utility routine that should not be called during normal Sigma operation.
static void	readWNversionMap(java.lang.String fileName, java.lang.String pattern, java.lang.String posNum) Read the version mapping files and store in the HashMap called "mappings".
static java.lang.String	removeTermPrefixes(java.lang.String formula)
static java.lang.String	rootFormOf(java.lang.String word)
static void	searchCoherence(java.lang.String fileWithPath) Take a file of tabtab and calculate the average Levenshtein distance for each ID.
static int	sensePOS(java.lang.String sense) Take a WordNet sense identifier, and return the integer part of speech code.
static boolean	sensoryOrMentalWord(java.lang.String word) test if a word is sensory or mental and return true if so
static java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>>	sensoryWords() Find all words associated with sensory, psychological and emotional concepts.
static void	showAllLeaves()
static void	showAllRoots()
static java.lang.String	subst(java.lang.String result, java.lang.String match, java.lang.String subst) A utility function that mimics the functionality of the perl substitution feature (s/match/replacement/).
static boolean	substTest(java.lang.String result, java.lang.String match, java.lang.String subst, java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>> hash) A utility function that mimics the functionality of the perl substitution feature (s/match/replacement/) but rather than returning the result of the substitution, just tests whether the result is a key in a hashtable.
static void	synesthesiaCompare(java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>> words, java.util.HashSet<java.lang.String> synwords) Compare Lievers list of synesthetic words with those derived from SUMO-WordNet
static java.util.HashSet<java.lang.String>	synestheticSynsets(java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>> words) Find all the words that exhibit links to multiple sensory modes in SUMO
static java.lang.String	synsetFromOntoNotes(java.lang.String onKey) Extract the nine digit synset ID corresponding to a word-POS.num sense key.
static java.lang.String	synsetToOneWord(java.lang.String s)
static void	testCommonParent() A method used only for testing.
static void	testIsValidKey() A method used only for testing.
static void	testSynesthesia()
static void	testWord()
static void	updateWNversion(java.lang.String path, java.lang.String versionPair) Port the mappings from one version of WordNet to another.
static void	updateWNversionProcess(java.lang.String fileName, java.lang.String pattern, java.lang.String posNum) This is a utility routine that should not be called during normal Sigma operation.
static void	updateWNversionReading(java.lang.String path, java.lang.String versionPair) Note that the "old" synset should be the second element of each line
static java.lang.String	verbPlural(java.lang.String verb) Return the plural form of the verb.
static java.util.HashSet<java.lang.String>	wordsToSynsets(java.lang.String word)
static void	writeTPTPWordNet(java.io.PrintWriter pw) Write TPTP format for WordNet

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

mappings

public static java.util.HashMap<java.lang.String,java.lang.String> mappings

POS-prefixed mappings from a new synset number to the old one.

TPTPidCounter

public static int TPTPidCounter

errorCount

public static int errorCount

WordNetRelations

protected static java.util.ArrayList<java.lang.String> WordNetRelations

withThoughtEmotion

public static boolean withThoughtEmotion

Constructor Detail

WordNetUtilities

public WordNetUtilities()

Method Detail

getBareSUMOTerm

public static java.lang.String getBareSUMOTerm(java.lang.String term)

Get a SUMO term minus its &% prefix and one character mapping suffix.

isValidSynset8

public static boolean isValidSynset8(java.lang.String synset)

Check whether a synset format is valid

isValidSynset9

public static boolean isValidSynset9(java.lang.String synset)

Check whether a synset format is valid

isValidKey

public static boolean isValidKey(java.lang.String senseKey)

Check whether a sense key format is valid

getPOSfromKey

public static java.lang.String getPOSfromKey(java.lang.String senseKey)

Extract the POS from a word_POS_num sense key. Should be an alpha key, such as "VB".

getWordFromKey

public static java.lang.String getWordFromKey(java.lang.String senseKey)

Extract the word from a word_POS_num sense key.

getSenseFromKey

public static java.lang.String getSenseFromKey(java.lang.String senseKey)

Extract the synset corresponding to a word_POS_num sense key.

getKeyFromSense

public static java.lang.String getKeyFromSense(java.lang.String synset)

Get the word_POS_num sense key corresponding to a 9 digit synset. Note that some adjective keys are listed as "adjuncts" with id '3' instead of '5' so we try that too in case of failure.

synsetFromOntoNotes

public static java.lang.String synsetFromOntoNotes(java.lang.String onKey)

Extract the nine digit synset ID corresponding to a word-POS.num sense key. see nlp.corpora.OntoNotes

removeTermPrefixes

public static java.lang.String removeTermPrefixes(java.lang.String formula)

convertTermList

public static java.util.ArrayList<java.lang.String> convertTermList(java.lang.String termList)

Convert a list of Terms in the format "&%term1 &%term2" to an ArrayList of bare term Strings

getSUMOMappingSuffix

public static char getSUMOMappingSuffix(java.lang.String term)

Get a SUMO term mapping suffix.

convertWordNetPointer

public static java.lang.String convertWordNetPointer(java.lang.String ptr)

posLetterToNumber

public static char posLetterToNumber(char POS)

posNumberToLetter

public static char posNumberToLetter(char POS)

posPennToNumber

public static char posPennToNumber(java.lang.String penn)

posNumberToLetters

public static java.lang.String posNumberToLetters(java.lang.String pos)

Convert a part of speech number to the two letter format used by the WordNet sense index code. Defaults to noun "NN".

posLettersToNumber

public static java.lang.String posLettersToNumber(java.lang.String pos)

Convert a part of speech number to the two letter format used by the WordNet sense index code. Defaults to noun "NN".

sensePOS

public static int sensePOS(java.lang.String sense)

Take a WordNet sense identifier, and return the integer part of speech code.

mappingCharToName

public static java.lang.String mappingCharToName(char mappingType)

subst

public static java.lang.String subst(java.lang.String result,
                                     java.lang.String match,
                                     java.lang.String subst)

A utility function that mimics the functionality of the perl substitution feature (s/match/replacement/). Note that only one replacement is made, not a global replacement.

Parameters:

result - is the string on which the substitution is performed.

match - is the substring to be found and replaced.

subst - is the string replacement for match.

Returns:

is a String containing the result of the substitution.

substTest

public static boolean substTest(java.lang.String result,
                                java.lang.String match,
                                java.lang.String subst,
                                java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>> hash)

A utility function that mimics the functionality of the perl substitution feature (s/match/replacement/) but rather than returning the result of the substitution, just tests whether the result is a key in a hashtable. Note that only one replacement is made, not a global replacement.

Parameters:

result - is the string on which the substitution is performed.

match - is the substring to be found and replaced.

subst - is the string replacement for match.

hash - is a hashtable to be checked against the result.

Returns:

is a boolean indicating whether the result of the substitution was found in the hashtable.

verbPlural

public static java.lang.String verbPlural(java.lang.String verb)

Return the plural form of the verb. Handle multi-word phrases to modify only the first word.

formatWords

public static java.lang.String formatWords(java.util.TreeMap<java.lang.String,java.lang.String> words,
                                           java.lang.String kbName)

HTML format a TreeMap of word senses and their associated synset

formatWordsList

public static java.lang.String formatWordsList(java.util.TreeMap<java.lang.String,java.util.ArrayList<java.lang.String>> words,
                                               java.lang.String kbName)

HTML format a TreeMap of ArrayLists word senses

mergeUpdates

public static void mergeUpdates()
                         throws java.io.IOException

Read in a file with a nine-digit synset number followed by a space and a SUMO term. If the term is more specific than the current mapping for that synset, replace the old term. This is a utility that is not normally called from the interactive Sigma system.

Throws:

java.io.IOException

processMissingLinks

public static void processMissingLinks(java.lang.String fileName,
                                       java.lang.String pattern,
                                       java.lang.String posNum)
                                throws java.io.IOException

This is a utility routine that should not be called during normal Sigma operation. It does most of the actual work for deduceMissingLinks()

Throws:

java.io.IOException

deduceMissingLinks

public static void deduceMissingLinks()
                               throws java.io.IOException

Use the WordNet hyper-/hypo-nym links to deduce a likely link for a SUMO term that has not yet been manually linked. This is a utility routine that should not be called during normal Sigma operation.

Throws:

java.io.IOException

updateWNversionProcess

public static void updateWNversionProcess(java.lang.String fileName,
                                          java.lang.String pattern,
                                          java.lang.String posNum)
                                   throws java.io.IOException

This is a utility routine that should not be called during normal Sigma operation. It does most of the actual work for updateWNversion(). The output is a set of WordNet data files with a "-new" suffix.

Throws:

java.io.IOException

readWNversionMap

public static void readWNversionMap(java.lang.String fileName,
                                    java.lang.String pattern,
                                    java.lang.String posNum)
                             throws java.io.IOException

Read the version mapping files and store in the HashMap called "mappings". Note that the "old" synset should be the second element of each line

Throws:

java.io.IOException

updateWNversionReading

public static void updateWNversionReading(java.lang.String path,
                                          java.lang.String versionPair)
                                   throws java.io.IOException

Note that the "old" synset should be the second element of each line

Throws:

java.io.IOException

updateWNversion

public static void updateWNversion(java.lang.String path,
                                   java.lang.String versionPair)
                            throws java.io.IOException

Port the mappings from one version of WordNet to another. It calls updateWNversionReading to do most of the work. It assumes that the mapping file has the new synset first and the old one second. File names are for the new WordNet version, which will need to have different names from the old version that WordNet.java needs to read in order to get the existing mappings. This is a utility which should not be called during normal Sigma operation. Mapping files are in a simple format produced by University of Catalonia and available at http://www.lsi.upc.edu/~nlp/web/index.php?option=com_content&task=view&id=21&Itemid=57 If that address changes you may also start at http://www.lsi.upc.edu/~nlp/web/ and go to Resources and then an item on WordNet mappings.

Throws:

java.io.IOException

numSynsets

public static int numSynsets(char pos)

Returns:

the number of synsets in WordNet for the given part of speech

printStatistics

public static java.lang.String printStatistics()

imageNetLinks

public void imageNetLinks()
                   throws java.io.IOException

Import links from www.image-net.org that are linked to WordNet and links them to SUMO terms when the synset has a directly equivalent SUMO term

Throws:

java.io.IOException

extractMeronyms

public static void extractMeronyms()

A utility to extract meronym relations as relations between SUMO terms. Filter out relations between genus and species, which shouldn't be meronyms

searchCoherence

public static void searchCoherence(java.lang.String fileWithPath)

Take a file of tabtab and calculate the average Levenshtein distance for each ID.

commentSentiment

public static void commentSentiment(java.lang.String fileWithPath)

writeTPTPWordNet

public static void writeTPTPWordNet(java.io.PrintWriter pw)
                             throws java.io.IOException

Write TPTP format for WordNet

Throws:

java.io.IOException

findLeavesInTree

public static java.util.HashSet<java.lang.String> findLeavesInTree(java.util.HashSet<java.lang.String> rels)

Find all the leaf nodes for a particular relation in WordNet. Note that the leaf must have a link from another node to be a leaf. No isolated nodes can be considered leaves.

Returns:

a list of POS-prefixed synsets

findPathsToRoot

public static java.util.ArrayList<java.util.ArrayList<java.lang.String>> findPathsToRoot(java.util.ArrayList<java.lang.String> base,
                                                                                         java.lang.String synset)

Find the complete path from a given synset. If multiple inheritance results in multiple paths, return them all.

lowestCommonParent

public static java.lang.String lowestCommonParent(java.lang.String s1,
                                                  java.lang.String s2)

findLeaves

public static java.util.HashSet<java.lang.String> findLeaves(java.lang.String rel)

Find all the leaf nodes for a particular relation in WordNet. Note that a node may be a leaf simply because it has no such link to another node.

Returns:

a list of POS-prefixed synsets

showAllLeaves

public static void showAllLeaves()

showAllRoots

public static void showAllRoots()

wordsToSynsets

public static java.util.HashSet<java.lang.String> wordsToSynsets(java.lang.String word)

Returns:

POS-prefixed synsets

synsetToOneWord

public static java.lang.String synsetToOneWord(java.lang.String s)

collapseSenses

public static java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>> collapseSenses()

Returns:

a Set of Sets where each interior Set consists of WordNet word senses that all map to a single SUMO term. The goal is to provide a way to collapse WordNet synsets that embody overly fine grained distinctions.

getAllHyponyms

public static java.util.HashSet<java.lang.String> getAllHyponyms(java.lang.String s)

Returns:

all the hyponyms of a given POS-prefixed synset

getAllHyponymsTransitive

public static java.util.HashSet<java.lang.String> getAllHyponymsTransitive(java.lang.String s)

Returns:

all the hyponyms of a given POS-prefixed synset

isHyponymousWord

public static boolean isHyponymousWord(java.lang.String word,
                                       java.util.HashSet<java.lang.String> synsets)

Returns:

whether the word is a possible hyponym of a given POS-prefixed synset

generateHyponymSets

public static void generateHyponymSets(java.lang.String filename)

Generate sets of all hyponymous words for each synset in a file

generateSUMOfromWNsubtree

public static void generateSUMOfromWNsubtree(java.lang.String synset,
                                             java.lang.String sumo)

Generate notional SUMO terms from WordNet

generateSUMOfromWN

public static void generateSUMOfromWN(java.lang.String synset,
                                      java.lang.String sumo)

Generate notional SUMO terms from WordNet

generateSUMOfromWN

public static void generateSUMOfromWN()

Generate notional SUMO terms from WordNet. Start with an equivalence Make each synset a notional SUMO term with its parent either the synset parent or the equivalence.

getSynsetsFromSUMO

public static java.util.ArrayList<java.lang.String> getSynsetsFromSUMO(java.lang.String sumo)

get all synsets corresponding to a SUMO term

getSynsetsFromSUMOList

public static java.util.HashSet<java.lang.String> getSynsetsFromSUMOList(java.util.Collection<java.lang.String> sumo)

get all synsets corresponding to a list of SUMO terms

getWordsFromSynsetList

public static java.util.HashSet<java.lang.String> getWordsFromSynsetList(java.util.Collection<java.lang.String> synsets)

get all words corresponding to a list of synsets

rootFormOf

public static java.lang.String rootFormOf(java.lang.String word)

sensoryWords

public static java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>> sensoryWords()

Find all words associated with sensory, psychological and emotional concepts. Return a set of words with String keys as to the human sense plus "emotion" and "thought"

synestheticSynsets

public static java.util.HashSet<java.lang.String> synestheticSynsets(java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>> words)

Find all the words that exhibit links to multiple sensory modes in SUMO

synesthesiaCompare

public static void synesthesiaCompare(java.util.HashMap<java.lang.String,java.util.HashSet<java.lang.String>> words,
                                      java.util.HashSet<java.lang.String> synwords)

Compare Lievers list of synesthetic words with those derived from SUMO-WordNet

testCommonParent

public static void testCommonParent()

A method used only for testing. It should not be called during normal operation.

testIsValidKey

public static void testIsValidKey()

A method used only for testing. It should not be called during normal operation.

sensoryOrMentalWord

public static boolean sensoryOrMentalWord(java.lang.String word)

test if a word is sensory or mental and return true if so

testWord

public static void testWord()

testSynesthesia

public static void testSynesthesia()

main

public static void main(java.lang.String[] args)

A main method, used only for testing. It should not be called during normal operation.