Wordnet and word similarity Lectures 11 and 12Centrality measures (for hw2)Lexical semantics: meaning of individual wordsWhat’s a word?Relationships between word meaningsHomonymyHomonymy causes problems for NLP applicationsPolysemySlide 9SynonymsButSynonymy is a relation between senses rather than wordsAntonymsHyponymyWordNetFormat of Wordnet EntriesSlide 17WordNet Noun RelationsWordNet Verb RelationsWordNet HierarchiesWord Sense Disambiguation (WSD)Two variants of WSD taskSupervised Machine Learning ApproachesSupervised WSD 1: WSD TagsWordNet BassInventory of sense tags for bassSupervised WSD 2: Get a corpusSupervised WSD 3: Extract feature vectorsTwo kinds of features in the vectorsExamplesSlide 31CollocationalBag-of-wordsCo-Occurrence ExampleNaïve Bayes TestDecision Lists: another popular methodLearning Decision ListsYarowskyWSD Evaluations and baselinesMost Frequent SenseCeilingBootstrappingSentences extracting using “fish” and “play”Slide 44WSD PerformanceWord similarityTwo classes of algorithmsThesaurus-based word similarityPath based similarityRefinements to path-based similarityProblem with basic path-based similarityInformation content similarity metricsInformation content similaritySlide 54Information content: definitionsResnik methodSlide 57Extended LeskSummary: thesaurus-based similarityProblems with thesaurus-based methodsDistributional methods for word similarityContext vectorIntuitionDistributional similarityDefining co-occurrence vectorsCo-occurrence vectors based on dependenciesMeasures of association with contextWeighting: Mutual InformationMutual information intuitionLin is a variant on PMISimilarity measures1Wordnet and word similarity Lectures 11 and 122Centrality measures (for hw2)How representative is a sentence of the overall content of a document–The more similar are sentence is to the document, the more representative it isjijiiSSsimKScentrality ),(1)(3Lexical semantics: meaning of individual wordsIntro to Lexical Semantics –Polysemy, Synonymy,…–Online resources: WordNetComputational Lexical Semantics–Word Sense DisambiguationSupervisedSemi-supervised–Word SimilarityThesaurus-basedDistributional4What’s a word?Definitions we’ve used: Types, tokens, stems, inflected forms, etc... Lexeme: An entry in a lexicon consisting of a pairing of a form with a single meaning representationA lemma or citation form is the grammatical form that is used to represent a lexeme.Ca rp et is the lemma for carpetsThe lemma bank has two senses:–Instead, a bank can hold the investments in a custodial account in the client’s name–But as agriculture burgeons on the east bank, the river will shrink even more.A sense is a discrete representation of one aspect of the meaning of a word5Relationships between word meaningsPolysemyHomonymySynonymyAntonymyHypernomyHyponomy6HomonymyLexemes that share a form–Phonological, orthographic or bothBut have unrelated, distinct meanings–Examplesbat (wooden stick-like thing) vs bat (flying scary mammal thing)bank (financial institution) versus bank (riverside)–Can be homophones, homographs, or both:Homophones:–Write and right–Piece and peace7Homonymy causes problems for NLP applicationsText-to-Speech–Same orthographic form but different phonological form bass vs bassInformation retrieval–Different meanings same orthographic formQUERY: bat careMachine TranslationSpeech recognition–Why?8PolysemyThe bank is constructed from red brickI withdrew the money from the bank Are those the same sense?–Which sense of bank is this?Is it distinct from (homonymous with) the river bank sense?How about the savings bank sense?9PolysemyA single lexeme with multiple related meanings (bank the building, bank the financial institution)Most non-rare words have multiple meanings–The number of meanings is related to its frequency–Verbs tend more to polysemy–Distinguishing polysemy from homonymy isn’t always easy (or necessary)10SynonymsWord that have the same meaning in some or all contexts.–filbert / hazelnut–couch / sofa–big / large–automobile / car–vomit / throw up–Water / H20Two lexemes are synonyms if they can be successfully substituted for each other in all situations–If so they have the same propositional meaning11ButThere are no examples of perfect synonymy–Why should that be? –Even if many aspects of meaning are identical–Still may not preserve the acceptability based on notions of politeness, slang, register, genre, etc.Example:–Water and H2012Synonymy is a relation between senses rather than wordsConsider the words big and largeAre they synonyms?–How big is that plane?–Would I be flying on a large or small plane?How about here:–Miss Nelson, for instance, became a kind of big sister to Benjamin.–?Miss Nelson, for instance, became a kind of large sister to Benjamin.Why?–big has a sense that means being older, or grown up–large lacks this sense13AntonymsSenses that are opposites with respect to one feature of their meaningOtherwise, they are very similar!–dark / light–short / long–hot / cold–up / down–in / outMore formally: antonyms can–define a binary opposition or at opposite ends of a scale (long/short, fast/slow)–Be reversives: rise/fall, up/down14Hyponymysuperordinate vehicle fruit furniture mammalhyponym car mango chair dogOne sense is a hyponym of another if the first sense is more specific, denoting a subclass of the other–car is a hyponym of vehicle–dog is a hyponym of animal–mango is a hyponym of fruitConversely–vehicle is a hypernym/superordinate of car–animal is a hypernym of dog–fruit is a hypernym of mango15WordNetA hierarchically organized lexical databaseOn-line thesaurus + aspects of a dictionaryVersions for other languages are under developmentAvr. noun has 1.23 senseAvr. verb has 2.16 senses Category EntriesNoun 117,097Verb 11,488Adjective 22,141Adverb 4,60116Format of Wordnet Entries17• The set of near-synonyms for a WordNet sense is called a synset (synonym set); it’s their version of a sense or a conceptExample: chump as a noun to mean ‘a person who is gullible and easy to take advantage of’• Each of these senses share this same gloss• Thus for WordNet, the meaning of this sense of chump is this list.18WordNet Noun Relations19WordNet