Natural Language Processing

Similar documents
Syntax Parsing 1. Grammars and parsing 2. Top-down and bottom-up parsing 3. Chart parsers 4. Bottom-up chart parsing 5. The Earley Algorithm

CS 598 Natural Language Processing

Constraining X-Bar: Theta Theory

Basic Syntax. Doug Arnold We review some basic grammatical ideas and terminology, and look at some common constructions in English.

Grammars & Parsing, Part 1:

11/29/2010. Statistical Parsing. Statistical Parsing. Simple PCFG for ATIS English. Syntactic Disambiguation

Basic Parsing with Context-Free Grammars. Some slides adapted from Julia Hirschberg and Dan Jurafsky 1

Inleiding Taalkunde. Docent: Paola Monachesi. Blok 4, 2001/ Syntax 2. 2 Phrases and constituent structure 2. 3 A minigrammar of Italian 3

Context Free Grammars. Many slides from Michael Collins

Proof Theory for Syntacticians

Introduction to HPSG. Introduction. Historical Overview. The HPSG architecture. Signature. Linguistic Objects. Descriptions.

Some Principles of Automated Natural Language Information Extraction

UNIVERSITY OF OSLO Department of Informatics. Dialog Act Recognition using Dependency Features. Master s thesis. Sindre Wetjen

Chapter 4: Valence & Agreement CSLI Publications

Ensemble Technique Utilization for Indonesian Dependency Parser

Approaches to control phenomena handout Obligatory control and morphological case: Icelandic and Basque

1/20 idea. We ll spend an extra hour on 1/21. based on assigned readings. so you ll be ready to discuss them in class

AQUA: An Ontology-Driven Question Answering System

Case government vs Case agreement: modelling Modern Greek case attraction phenomena in LFG

Compositional Semantics

Enhancing Unlexicalized Parsing Performance using a Wide Coverage Lexicon, Fuzzy Tag-set Mapping, and EM-HMM-based Lexical Probabilities

ENGBG1 ENGBL1 Campus Linguistics. Meeting 2. Chapter 7 (Morphology) and chapter 9 (Syntax) Pia Sundqvist

Universal Grammar 2. Universal Grammar 1. Forms and functions 1. Universal Grammar 3. Conceptual and surface structure of complex clauses

Ch VI- SENTENCE PATTERNS.

Construction Grammar. University of Jena.

Underlying and Surface Grammatical Relations in Greek consider

Informatics 2A: Language Complexity and the. Inf2A: Chomsky Hierarchy

Developing Grammar in Context

A Minimalist Approach to Code-Switching. In the field of linguistics, the topic of bilingualism is a broad one. There are many

Accurate Unlexicalized Parsing for Modern Hebrew

A Graph Based Authorship Identification Approach

Project in the framework of the AIM-WEST project Annotation of MWEs for translation

Natural Language Processing. George Konidaris

Argument structure and theta roles

The presence of interpretable but ungrammatical sentences corresponds to mismatches between interpretive and productive parsing.

Minimalism is the name of the predominant approach in generative linguistics today. It was first

THE INTERNATIONAL JOURNAL OF HUMANITIES & SOCIAL STUDIES

Intensive English Program Southwest College

Developing a TT-MCTAG for German with an RCG-based Parser

Words come in categories

Achim Stein: Diachronic Corpora Aston Corpus Summer School 2011

LTAG-spinal and the Treebank

Parsing of part-of-speech tagged Assamese Texts

Today we examine the distribution of infinitival clauses, which can be

Update on Soar-based language processing

Mercer County Schools

ELD CELDT 5 EDGE Level C Curriculum Guide LANGUAGE DEVELOPMENT VOCABULARY COMMON WRITING PROJECT. ToolKit

Pseudo-Passives as Adjectival Passives

The Interface between Phrasal and Functional Constraints

Emmaus Lutheran School English Language Arts Curriculum

Semantic Inference at the Lexical-Syntactic Level for Textual Entailment Recognition

LNGT0101 Introduction to Linguistics

The College Board Redesigned SAT Grade 12

Objectives. Chapter 2: The Representation of Knowledge. Expert Systems: Principles and Programming, Fourth Edition

Reading Grammar Section and Lesson Writing Chapter and Lesson Identify a purpose for reading W1-LO; W2- LO; W3- LO; W4- LO; W5-

Content Language Objectives (CLOs) August 2012, H. Butts & G. De Anda

The Effect of Multiple Grammatical Errors on Processing Non-Native Writing

Chunk Parsing for Base Noun Phrases using Regular Expressions. Let s first let the variable s0 be the sentence tree of the first sentence.

Hyperedge Replacement and Nonprojective Dependency Structures

What the National Curriculum requires in reading at Y5 and Y6

Specifying a shallow grammatical for parsing purposes

CAS LX 522 Syntax I. Long-distance wh-movement. Long distance wh-movement. Islands. Islands. Locality. NP Sea. NP Sea

"f TOPIC =T COMP COMP... OBJ

The stages of event extraction

Analysis of Probabilistic Parsing in NLP

SAMPLE. Chapter 1: Background. A. Basic Introduction. B. Why It s Important to Teach/Learn Grammar in the First Place

Construction Grammar. Laura A. Michaelis.

Using dialogue context to improve parsing performance in dialogue systems

Sample Goals and Benchmarks

Prediction of Maximal Projection for Semantic Role Labeling

Dependency Annotation of Coordination for Learner Language

Parsing natural language

Chapter 1 The functional approach to language and the typological approach to grammar

Adjectives tell you more about a noun (for example: the red dress ).

Loughton School s curriculum evening. 28 th February 2017

Target Language Preposition Selection an Experiment with Transformation-Based Learning and Aligned Bilingual Data

Hindi Aspectual Verb Complexes

Copyright and moral rights for this thesis are retained by the author

Annotation Projection for Discourse Connectives

An Interactive Intelligent Language Tutor Over The Internet

Character Stream Parsing of Mixed-lingual Text

Towards a Machine-Learning Architecture for Lexical Functional Grammar Parsing. Grzegorz Chrupa la

What Can Neural Networks Teach us about Language? Graham Neubig a2-dlearn 11/18/2017

Cross-lingual Transfer Parsing for Low-Resourced Languages: An Irish Case Study

cambridge occasional papers in linguistics Volume 8, Article 3: 41 55, 2015 ISSN

cmp-lg/ Jul 1995

Adapting Stochastic Output for Rule-Based Semantics

Control and Boundedness

Hindi-Urdu Phrase Structure Annotation

Derivational: Inflectional: In a fit of rage the soldiers attacked them both that week, but lost the fight.

A Computational Evaluation of Case-Assignment Algorithms

On the Notion Determiner

Writing a composition

Psychology and Language

The Pennsylvania State University. The Graduate School. College of the Liberal Arts THE TEACHABILITY HYPOTHESIS AND CONCEPT-BASED INSTRUCTION

Machine Learning from Garden Path Sentences: The Application of Computational Linguistics

Linguistic Variation across Sports Category of Press Reportage from British Newspapers: a Diachronic Multidimensional Analysis

Aspectual Classes of Verb Phrases

Feature-Based Grammar

The Strong Minimalist Thesis and Bounded Optimality

Transcription:

McDonald et al. 2006 Natural Language Processing Info 159/259 Lecture 16: Dependency syntax (Oct 18, 2018) David Bamman, UC Berkeley

Context-free grammar A context-free grammar defines how symbols in a language combine to form valid structures NP Det Nominal NP ProperNoun non-terminals Nominal Noun Nominal Noun Det a the Noun flight lexicon/ terminals

Constituents Every internal node is a phrase my pajamas in my pajamas elephant in my pajamas an elephant in my pajamas shot an elephant in my pajamas I shot an elephant in my pajamas Each phrase could be replaced by another of the same type of constituent

Formalisms Phrase structure grammar (Chomsky 1957) Dependency grammar (Mel čuk 1988; Tesnière 1959; Pāṇini) last week today

Dependency syntax Sgall, Dependency-based formal description of language (1994) Mel čuk, Dependency Syntax: Theory and Practice (1988) Tesnière, Éléments de syntaxe structurale (1959) Medieval theories of grammar (Covington 1984) Pānini grammar of Sanskrit (ca. 5th-century BCE)

Dependency syntax Sentence diagramming

Dependency syntax Between the word and its neighbors, the mind perceives connections, the totality of which forms the structure of the sentence. The structural connections establish dependency relations between the words. Each connection in principle unites a superior and an inferior term. Tesnier 1959; Nivre 2005

Dependency syntax Dependency syntax doesn t have non-terminal structure like a CFG; words are directly linked to each other.

Dependency syntax Syntactic structure = asymmetric, binary relations between words. Tesnier 1959; Nivre 2005

Dependency How do we decide which of a pair of words is the head and which is the dependent?

Dependency Many (conflicting) frameworks: Head determines the syntactic category of a construction Head is obligatory; dependents are optional Head selects dependents and determines whether the dependent is required The form of the dependent depends on the head (e.g., agreement between nouns/verbs, adjectives/ nouns) The linear position of a dependent is specified with respect to the head.

Coordination Nivre 2005

Case marking prepositions Nivre 2005

Trees A dependency structure is a directed graph G = (V,A) consisting of a set of vertices V and arcs A between them. Typically constrained to form a tree: Single root vertex with no incoming arcs Every vertex has exactly one incoming arc except root (single head constraint) There is a unique path from the root to each vertex in V (acyclic constraint)

Trees Unlike phrase-structure trees, dependency trees aren t tied to the linear order of the words in a sentence. Adding a constraint derived from the linear order of words in a sentence allows for more efficient parsing algorithms (as we ll see on Tuesday).

Word order Dependency relations belong to the structural order of a sentence, not the linear order. This is different from a phrase-structure tree, where the syntax is constrained by the linear order of the sentence (a different linear order yields a different parse tree). Tesnière 1959

Free word order

Free word order

Projectivity An arc between a head and dependent is projective if there is a path from the head to every word between the head and dependent.

Dependencies vs constituents Dependency links are closer to semantic relationships; no need to infer the relationships from the structure of a tree A dependency tree contains one edge for each word, no intermediate hidden structures that also need to be learned for parsing. Easier to represent languages with free word order. Covington 2001

S NP VP V NP PP P NP noun verb noun prep noun NBC suspended Williams on Tuesday

S Who did what to whom? NP VP V NP PP P NP noun verb noun prep noun NBC suspended Williams on Tuesday subject: S NP VP direct object: S NP (VP NP )

S NP VP V NP PP P NP noun verb noun prep noun NBC suspended Williams on Tuesday dobj nsubj prep pobj

Dependency grammar Captures binary relations between words nsubj(nbc, suspended) dobj(williams, suspended) NBC suspended Williams on Tuesday dobj nsubj prep pobj

Data police found five.030 bullets 1 police found seven dead rebels 3 police found two hidden cameras 2 police found wanders lover 1 police found 211 pounds 4 NELL SVO triples (604 million nsubj+dobj relations from 230B words on the web police found Marcia 3 police found bank draft 1 police found diskette 2 police found five marijuana plants 3 police found items used 1 police found judge 5

Dependency- Based Word Embeddings Levy & Goldberg, ACL 2014 http://irsrv2.cs.biu.ac.il:9998/

Universal Dependencies http://universaldependencies.org

Universal Dependencies Developing cross-linguistically consistent treebank annotation for many languages Goals: Facilitating multilingual parser development Cross-lingual learning Parsing research from a language typology perspective. http://universaldependencies.org

Universal Dependencies http://universaldependencies.org

UD Principles Dependency relations mainly hold between content words. http://universaldependencies.org/u/overview/syntax.html

UD Principles Function words dependent on closest related content word http://universaldependencies.org/u/overview/syntax.html

UD Principles Punctuation typically depends on phrasal head http://universaldependencies.org/u/overview/syntax.html

UD Principles http://universaldependencies.org/u/overview/syntax.html

nsubj Syntactic subject of active verbs

nsubj When the verb is copular (nominal/adjective predicatives like she is my boss ), the subject depends on the complement.

Expletives nsubj

nsubj Relative clauses with finite verbs also often have syntactic subjects expressed.

nsubj:pass Syntactic subject of passive verbs

obj Generally, the entity that is acted upon as the direct object of the predicate.

iobj Indirect object: recipients of ditransitive verbs of exchange (verbs requiring two objects) nsubj iobj obj She teaches her daughters math She told her daughtesr a story

obl Any nominal functioning as non-required argument or adjunct of a verb, including temporal and locational nominal modifiers and agents of passive verbs

ccomp Clausal complements, including dialogue

advcl A clause that modifies another predicate (temporal clauses, consequence, conditional clauses, purpose clauses)

acl Clausal modifier of a nominal, including relative clauses. The head verb of the embedded clause modifies the nominal it qualifies

acl:relcl Clausal modifier of a nominal, including relative clauses. The head verb of the embedded clause modifies the nominal it qualifies

conj The elements that are coordinated; the head is the first conjunct

Intransitive verbs

Transitive verbs

Ditransitive verbs

Nonverbal predication

Nonverbal predication Equation (aka identification): she is my mother Attribution: she is nice Location: she is in the bathroom Possession: the book is hers Benefaction: the book is for her Existence: there is food (in the kitchen)

Oblique nominals

Vocatives

Aux aux adds tense, aspect, mood, voice or evidentiality

cop cop links a non-verbal predicate to subject

Open the pod bay doors, HAL [David Bowman, 2001] Nobody puts baby in a corner [Johnny Castle, Dirty Dancing]

Tuesday Algorithms for dependency parsing Shift-reduce parsing Graph-based parsing