nlp. introduction to nlp what is the meaning of cat? its pronunciation? part of speech? what is the...
Post on 05-Jan-2016
Embed Size (px)
NLPIntroduction to NLPMorphology and the LexiconMental LexiconWhat is the meaning of cat? Its pronunciation? Part of speech?What is the meaning of wug?What is the meaning of cluvious?Compare traftful and traftless?Morphology of these wordsIntuition and productivity RunsAllomorphs cats/oxen, played/swungAffixes
Derivational MorphologyEr (many examples)What do these morphemes mean?Ness, able, ing, re, un, er (adj)JJ V + -ableRecursion: unconcernednessesAmbiguous uncloggableNot ambiguous unbelievable why?
JJ-ableVdrinkAnswer to the QuizUncloggableunable to be cloggedable to be uncloggedUnbelievableunable to be believed? able to be unbelieved
Morphological ExamplesReduplicationamigo = friend, amimgo = friends (in Pangasinan) [Rubino 2001]savali = he travels, savavali = they travel (in Samoan)Templatic morphology (e.g., Semitic languages): lmd (learn), lamad (he studied), limed (he taught), lumad (he was taught)Circumfixesspielen gespielt (in German)Pig LatinappyhayVerlancfran, ripou (from lenvers, Franais, pourri)Massa-freakin-chusettswhere can you insert freakin in education?Answer to the QuizThe freakin infix is inserted to the left of the syllable that bears the main stressedu-freakin-cation * educa-freakin-tion * e-freakin-ducationthough there can be exceptions
Inflectional MorphologyTense, number, person, mood, aspectFive verb forms in English40+ forms in FrenchSix cases in Russian:http://www.departments.bucknell.edu/russian/language/case.htmlUp to 40,000 forms in Turkish (you cause X to cause Y to do Z)8Morphological Analysissleeps = sleep + V + 3P + SGdone = do + V + PP9Check provenance!!Turkish Vowel HarmonyBack vowelsin the room odadaat the door kapdaFront vowelsat home evdeat the lake gldeon the bridge kprdeFrontBackUnroundedRoundedUnroundedRoundedHighiuLoweaohttp://webspace.ship.edu/cgboer/morphology.html
10NACLO ProblemTurkishwww.naclo.cs.cmu.edu/problems2010/F.pdf NACLO SolutionTurkishwww.naclo.cs.cmu.edu/problems2010/FS.pdf
Slide from Kemal OflazerAgglutinative LanguagesUsed with permission13amefutoAme(rican) Foot(ball)aisu kurmuice creamaidoruidolaptoapartmentbaikubikebariafurbarrier freekonpytcomputerdesukudesk (at a news agency)doramadrama (on TV)erebtelevatoresukartescalatorfuraidopotetofried potato (French fries)gurasuglass (for drinking)happendohappy end(ing)hottokkihotcake (pancake)kash nattsucashew nutkhcoffeekurabuclubkbdokeyboardkyanpncampaignkyappucappsokonperso(nal) com(puter)psonaru konpytpersonal computerrejumeresumeresutoranrestaurantrimokonremo(te) con(trol)saradasaladtabakotobaccoterebigmutelevision gamezeminruseminarMOVE14Introduction to NLPOther Levels of Linguistic AnalysisSemanticsSemanticsLexical semantics and compositional semanticsLexical SemanticsHypernyms, hyponyms, antonyms, meronyms and holonyms (part-whole relationship, tire is a meronym of car), synonyms, homonymsSenses of words, polysemous wordsCollocationswhite hair, white wineIdiomsto kick the bucketCompositional SemanticsHow to understand the meaning of a sentence based on the meaning of its components.PragmaticsThe study of how knowledge about the world and language conventions interact with literal meaning.Speech actsResolution of anaphoric relationsModeling of speech acts in dialogue
Other AreasSociolinguisticsinteractions of social organization and language.Historical linguisticschange over time.Linguistic typologyLanguage acquisitionL1 and L2Psycholinguisticsreal-time production and perception of languageNLP