3. Linguistic Essential

1
3. Linguistic Essential

• ???? ???
• ???

2
3. Linguistic Essential

• 3.1 Parts of Speech and Morphology
• Nouns and pronouns
• Determiners and adjectives
• Verbs
• Other parts of speech (adverbs, prepositions,
  particles)
• 3.2 Phrase Structure
• Phrase structure grammars
• Dependency Arguments and adjuncts
• X' theory
• Phrase structure ambiguity
• 3.3 Semantics and Pragmatics

3
3.1 Parts of Speech and Morphology (1-1)

• Parts of speech(POS) Syntactic or grammatical
  categories
• show similar syntactic behavior
• Three important parts of speech
• Nouns (refer to people, animals, concepts and
  things)
• Verb (used express the action in the sentence)
• Adjectives (describe properties of nouns)
• Substitution test most basic test for words
  belonging to the same class

Children eat sweet candy
sad intelligent The green one
is in the corner fat
4
3.1 Parts of Speech and Morphology (1-2)

• Many words have multiple parts of speech
• Word classes
• 1. open (or lexical) categories
• nouns, verbs, adjectives
• large number of members
• new words added
• 2. closed (or functional) categories
• prepositions (of, on), determiners(the, a)
• a few members
• normally have a clear grammatical use
• Various parts of speech for a word listed in a
  lexicon
• Systems of parts of speech
• Tradition about 8 categories
• Corpus linguists sets of abbreviations for
  naming word classes ( POS tags )

Too much boiling will candy the molasses Have
a candy from the box
5
3.1 Parts of Speech and Morphology (2-1)

• Word categories systematically related by
  Morphological processes
• Formation of the plural form from the singular
  form of the noun
• Dog ? Dog-s
• Morphology
• Very important in NLP
• Language is productive
• ??? ??
• ? ??? ?? ?? ?? ??? ??
• ??? ??? ???? ??? ??
• ?? ???? ???? ??? ?? ??

6
3.1 Parts of Speech and Morphology (2-2-1)

• Types of Morphological processes
• 1. Inflexion
• Systematic modifications of a root for by means
  of prefixes and suffixes
• indicatite grammatical distinction (singular/
  plural)
• Varies features (tense, number, plurality)
• No change word class or meaning significantly
• Inflectional forms of a word manifestations of
  a single lexeme

7
3.1 Parts of Speech and Morphology (2-2-2)

• 2. Derivation
• Less systematic (wide ? widely/ old ? oldly/
  difficult ? difficultly)
• More radical change of syntactic category
• Involves a change in meaning
• Suffixes of derivation -en(weak-en)/
  -able(accept-able)/ -er(teach-er)

wide(adj) widely(adv) a wide river
?? ?? ???? It is widely believed ?? ???? ???
???
derivation
8
3.1 Parts of Speech and Morphology (2-2-3)

• 3. Compounding two or more words into a new word
• Noun-noun compounds combinations of two other
  nouns
• written as separate words,
• pronounced as a single word,
• Denote a single semantic concept (?? lexicon? ??)
• Other compounds involving adjectives, verbs,
  prepositions

tea kettle, disk driver, college degree
down market, (to)overtake, mad cow disease
9
3.1.1 Noun and pronouns (1)

• Entities in the world people, animals, things
• Number ex) English (?????? ??? ??)
• Gender

Type of inflection Instances number singular,
plural gender feminine, masculine,
neuter case nominative, genitive, dative,
accusative
plural form suffix-s regular dog
dogss personpersonsz speechspeeches?s Ir
regular child children womanwomen
English He, she, it Latin -a (fem) -us
(mas) fili-us (son male child) fili-a
(daughter female child) German mädchen
(daughter neuter arbitrary)
10
3.1.1 Noun and pronouns (2)

• Cases nouns appears in different forms when they
  have different functions (subject, object, etc.)
  in the sentence

Latin Filius (subject) filium (object) English
no real case inflections only case relationship
? genitive(describes the possessor) s,
(after words ending in s)
CLITIC phrasal affix - womens house - The
person you mets house was broken into
11
3.1.1 Noun and pronouns (3)

• Pronoun variables in that they refer to a person
  or thing that is somehow salient in the discourse
  context
• Only words in English which appear in different
  forms when they are used as the subject and the
  object of the sentence
• personal pronouns
• Possessive pronouns my car, a friend of mine
• Reflexive pronouns always refer to nearby
  antecedent in the same sentence, normally the
  subject of the sentence

After Mary arrived in the village, she looked for
a bed-and-breakfast
Nominative Subject case
Accusative Object case
Anaphors refer to something very nearby in the
text
12
3.1.1 Noun and pronouns (4)
NN (singular nouns) NNP (proper nouns ex.
Mary, Korea) NR (adverbial nouns ex. west
tomorrow) NNS (plural nouns) NNPS (plurar
proper nouns) NRS (plural adverbial nouns) NN
(possessive singular nouns) NNS (possessive
plural nouns) NNP (possessive singular proper
nouns) NNPS (possessive plural proper
nouns) NR (possessive adverbial nouns)
Nominative PPS (3SG) PPSS (1SG, 2SG,
PL) Accusative PPO Possessive PP / PP (2nd
possessive) Reflexive PPL / PPLS (plural)
13
3.1.2 Words that accompany nouns (1)

• Determiners(???) and adjectives(???)
• Determiner describe the particular reference of
  a noun
• Article (the a(an))
• Demonstratives (this, that)
• Adjectives describe properties of noun
• Attributive (Adnominal)
• Predicative (complement of be)
• Agreement adjective( article) agree with the
  noun (case, number, gender)
• Comparative/ Superlative (gt Positive (rich))
• -er/ -est (richer/ richest)
• periphrastic forms (more intelligent/ most
  intelligent )
• Quantifiers (???)
• Pre-quantifier (all, many)
• Nominal pronoun (one ,something, anything)
• Interrogative pronouns/ determiners with or
  instead of nouns

14
3.1.2 Words that accompany nouns (2)
JJ (positives adj) JJR (comparatives adj) JJT
(superlatives adj) JJS (semantically
superlatives adj chief) NUMBERS CD (cardinals
one, two) OD (ordinals first, second)
AT (articles) DT (singular determiners this,
that) DTS (plural determiners these,
those) DTI (? ?? ?? ?? some, any) DTX (double
conjunction either, neither)
WDT (wh-determiner what, which) WP (possessive
wh-pronoun whose) WPO (objective wh-pronoun
whom, which, that) WPS (nominative wh-pronoun
who, which, that)
ABN (pre-quantifier all, many) PN (nominal
pronoun one, something) EX (??? ?? ???? ??
there)
15
3.1.3 Verbs

• Describe action
• Morphological forms of regular verbs

VB Base form take
VBD Past tense took
VBG Gerund present participle taking
VBN Past participle taken
MD Modal auxiliaries can, may, must, could, might
MD see table 4.6 ! be, have, do see table 4.6 ! be, have, do
16
3.1.4 Other parts of speech

• Adverbs
• Modify a verb (specify place, time, manner,
  degree)
• Modify adjectives and adverbs
• Preposition
• Express spatial relationships
• Particle(???lt??? ! suffixgt)
• Subclass of prepositions most prepositions do
  double duty as particles
• Construct phrasal verb entering into strong
  bounds with verbs
• Separate lexical entry (? different syntactic
  semantic properties)
• preposition ? particle
• Conjunctions complementizers

She ran up a hill She ran up a bill
17
3.1.4 Other parts of speech
RB Ordinary adverb simply, late, well, little
RBR Comparative adverb later, better, less
RBT Superlative adverb latest, best, least
QL Qualifier degree adverb very, too, extremely (???? ???? ??)
QLP Post-qualifier enough, indeed
WQL Wh-qualifier how
WRB Wh-adverb how, when, where
IN Prepositions
RP Particles
CC Conjunctions and, or, but
CS Subornating conjunctions (? Complementizers that) that, because, if, before,
18
3.2 Phrase structure(???) (1)

• Syntax study of the regularities and constraints
  of word order and phrase structure
• Constituents(????) 3.2.1 Phrase structure
  grammars (1)
• able to occur various positions
• uniform syntactic possibilities for expansion
• Paradigmatic relationship all element that can
  be replaced for each other in certain syntactic
  position are member of one paradigm.
• Syntagmatic relationship ??? ?(phrase
  (syntagma))? ??? ? ??? ??(?) ??? ??

She The woman The tall woman The very tall woman The tall woman with sad eyes saw him the man the short man the very short man the short man with red hair .
19
3.2 Phrase structure(???) (2)

• Typical English phrase structure
• Rewrite rule

S NP VP That man VBD NP
PP caught the butterfly IN
NP with a net
20
3.2 Phrase structure(???) (3)

• Noun phrases (NP)
• Head noun
• Arguments of verb
• Prepositional phrases (PPs)
• Head preposition
• Contain a noun phrase complement
• Express spatial, temporal locations, etc.
• Verb phrases (VP)
• Head verb
• Organize all elements of the sentence
• Adjective phrases (APs) very sure of herself,
  quite certain to succeed

(determiner) (adjective phrase) noun
(post-modifier) (optional) (head)
prepositional phrases, clausal modifiers
21
3.2.1 Phrase structure grammars (1)

• Word order
• Change in meaning English
• Do not change in meaning Latin ( free word
  order language)
• Declaratives, Interrogatives (inversion),
  Imperatives
• Rewrite rules (constituency? ???1) 3.2 Phrase
  structure (1)
• used to generate sentences
• A ? B, A ? B C
• Generation

S ? NP VP AT ? the AT NNS chidren NP ?
AT NN NNS ? students NP PP mountains VP
PP slept VP ? VBD VBD ? ate VBD
NP saw etc.

• S
• NP VP
• AT NNS VBD
• The children slept

22
3.2.1 Phrase structure grammars (2)

• Tree (constituency? ??? 2)
• Terminal nodes???? ??? ?? ??
• Non terminal nodes (Internal nodes) ??? ??, ?
• The order of daughters generates the word order
  of sentence
• Bracketing (constituency? ??? 3) grouping

S NP VP AT NNS VBD The children
slept
S NP AT THE NNS children VP VBD ate NP
AT the NN cake
23
3.2.1 Phrase structure grammars (3)

• Recursivity recursive expansions
• A property of most formalizations of natural
  language syntax in terms of rewrite rules
• The fact that there are constellations in which
  rewrite rules can be applied an arbitrary number
  of times
• Non-local dependencies challenge some
  Statistical NLP approach
• Syntactically dependent even though they occur
  far apart in a sentence
• Subject-verb agreement(number, person)
• Long-distance dependencies
• Wh-extraction
• Empty nodes ?, e ex. NP? ?

S NP S Which book MD S
Should NP VP peter VB NP
buy e
Should Peter buy a book? Which book should Peter
buy?
24
3.2.2 Dependency Arguments and adjuncts (1)

• Dependency ( dependency grammar ????) ?? ?????
  ?? ????? ???? ????? ??
• ?? ??? ?? ????(?? ??)? ???? ??
• ??? ?? ?? ??? ??? ??? ??? ?? ???? ?? ?? ???, ? ?
  ?? ??? ???? ?? ?? ???? ?? ?? ???? ??? ??
• ?? ?? ? ?? ?? ??? ?? ??(binary relation)
• ?? ??(governor head) ?? ??(dependent)
• ??(????)? ?? ??
• ??? ???? ?? ???(phrase structure)? ??? ????? ??
• ??? ????, ??? ??? ???? ??? ?? (? ???? ???)
• ????? ?? ???? ?? ??
• ??? ??? ???? ???? ???? ??? ??? ??? ????? ?? ??
  ??, ?? ??? ?? ??? ???? ??
• -??, ????, etc.
• ??? ?? ?? ?? ??
• -???, etc.
• ??? (??? ????)

I put a pen on the table. ?? ?? ?? ?? ???. A pen
put I on the table. ?? ?? ?? ?? ???. I put a
pen. ?? ?? ???.
25
3.2.2 Dependency Arguments and adjuncts (2)

• Arguments (dependent) of verbs
• express entities that are centrally involved in
  the activity of the verb
• NPs, PPs, VPs
• Semantic roles
• Agent(???)/ Patient(???)
• Grammatical relations
• Subject
• Object
• direct object (patient)
• indirect object (recipients(???)) prepositional
  phrase
• Roles(relations) change by voice alternations
• Active Passive
• English patient ? subject / agent ? oblique
  role (by-phrase)
• Others change in case marking/ morphology on the
  verb

26
3.2.2 Dependency Arguments and Adjuncts (3)

• Adjunct (dependent)
• Less tight link to the verb
• Always optional (many complement obligatory)
• Move around more easily than complement
• Phrases describing time, place, manner of action,
  or state
• yesterday, in Paris, with great interest
• Difficult to distinguish adjunct and compliment
• Intermediate degree of selection ?
• Statistical NLP degree of association between
  a verb and a dependent

He put the book on the table (obligatory). He
gave his presentation on the stage (optional).
He will retire in Florida.
27
3.2.2 Dependency Arguments and adjuncts (4)

• Subcategorization(?????)
• A verb subcategorizes for a particular complement
• ex) bring subcategorizes for an object
• Subcategorized arguments
• Subject, Object, Prepositional phrase,
• Predicative adjective, Bare infinitive,
  Infinitive with to,
• Participial phrase, That-clause, Question form
  clauses
• SS Bar constituent relative clause, main
  clause questions

28
3.2.2 Dependency Arguments and adjuncts (5)

• Syntactic regularities about complements
• Subcategorization frame patterns of arguments
• A particular set of arguments that a verb can
  appear with
• Semantic regularities between constituents
• Selectional restrictions (preferences)
• bark (dogs as subjects)/ eat (edible as objects)
• Violation of selectional preferences odd
  sentence

Intransitive verb NPsubject
Transitive verb NPsubject, NPobject
Ditransitive verb NPsubject, NPdirect object, NP indirect object
Intransitive with PP NPsubject, PP
Transitive with PP NPsubject, NPobject, PP
Sentential comp NPsubject, clause
Transitive with sentential comp NPsubject, NPobject, clause
29
3.2.3 X theory

• Phrases structure rules as presented above do not
  predict any systematicity in the way that phrases
  in natural languages are made, nor any
  regularities for the appearance of different
  kinds of dependents in clauses.
• Head of a phrase a word
• A broad systematicity in the way dependents
  arrange themselves around a head in a phrase
  head/complements

NP Det N The AP
N definitive N PP study of
subcategorization
(X) N (XP) N Basic 2 level Can have more or
fewer level
30
3.2.4 Phrase structure ambiguity (1)

• rewrite rules used in Parsing
• Parse phrase structure tree that is constructed
  from a sentence
• Phrase structure ambiguity (syntactic structure
  ambiguity)
• ex) 100 parse for a English sentence
• Attachment ambiguity
• Phrase that could have been generated by 2
  different nodes
• Different attachments have different meanings.

The children ate the cake with a
spoon. Attachment to the verb phrase
(instrument) Attachment to the noun phrase
(which cake was eaten)
31
3.2.4 Phrase structure ambiguity (2)

• S
• NP VP
• AT NNS VP PP
• The children VBD NP IN NP
• ate AT NN with AT NN the cak
  e a spoon

S NP VP AT NNS VBD NP The
children ate NP PP AT NN IN
NP the cake with AT NN
a spoon
32
3.2.4 Phrase structure ambiguity (3)

• Garden paths(??? ???)
• Additional words in the sentence that do not seem
  to belong there
• ? adopt a spurious parse
• ? backtrack to try to construct the right parse
• Rarely problem in spoken language( intonational
  patterns, pause )
• No path at all (not covered by the grammar)
• Syntactic illformedness (ungrammatical) no
  interpretation
• Semantic abnormality semantic, pragmatic,
  cultural oddness

The horse ran past the barn fell ? The horse
fell after it had been raced past the barn
Slept chidren the.
Coloress green ideas sleep furiously.
33
3.3 Semantics and Pragmatics (1)

• Semantics study of the meaning of words,
  constructions, and utterances
• 1. Lexical semantics
• Lexical hierarchy
• Ambiguity refer to homonymy polysemy
• Hypernymy Animal(general) is hypernym of
  cat(specialized)
• Antonyms words with opposite meaning hot /
  cold , long / short
• Meronymy part-whole relationship
• Meronym(holonym) tire (car), leaf (tree)
• Synonyms words with the same or very similar
  meaning car / automobile
• Homonyms different words that are written the
  same way bank
• Polyseme words meaning are related branch
• Homophony written the same way, identical
  pronunciation bass(???, ??)

34
3.3 Semantics and Pragmatics (2)

• 2. Study of how meanings of individual words are
  combined into the meaning of
• sentences (gtgt discourses)
• Compositionality the meaning of the whole can be
  predicted from the meaning of the parts
• white paper(white), white hair(grey), white
  skin(rose), white wine(yellow)
• Collocations
• meaning of the whole sum of the meaning of the
  part some additional semantic component that
  cannot be predicted from the parts
• Idiom relationship between the meaning of the
  words and the meaning of the phrase is completely
  opaque
• To kick the bucket ( die)
• Scope quantifier have a scope which extends over
  one or more phrases or clauses
• Everyone didnt go to the movie.

35
3.3 Semantics and Pragmatics (3)

• Discourse analysis
• Relationships between sentences in the text
• Part of pragmatics
• Pragmatics study of how knowledge about the
  world and language conventions interact with
  literal meaning
• Anaphoric relations important for information
  extraction

Mary helped Peter get out of the cab. He thanked
her Mary helped the other passenger out of the
cab. The man had asked her to help him because
of his foot injury.
Which Hurricane caused more than a billion
dollars worth of damage ? need pragmatic
information