Incremental parsing and ambiguous sentences - each incoming materials are attached immediately - a single analysis is pursed - given the uncertainty,

Slides:



Advertisements
Similar presentations
Gerunds & Infinitives Verb – ing To + base verb.
Advertisements

TEACHING GRAMMAR Bui Thi Thao Truong Thuy Duong.
The art of making questions part II. decide / you / do/ to / Why / again/ did / it The answer: Why did you decide to do it again? Task one. Make these.
Intro to NLP - J. Eisner1 Modeling Grammaticality [mostly a blackboard lecture]
STAGES OF COMPREHENSION discourse modelling semantic analysis syntactic “parsing” lexical access phonemic analysis sensory processing.
Phrasal verbs 10/11/2011Ana Suárez. Starter unit Go out Have a romantic relationship Split up Break up Ask out Invite someone on a date 10/11/2011Ana.
Analysing Syntax 1 Lesson 8B.
Morphology and Syntax Practice 1 Lesson 9A The top of the tree So the tree for any sentence (S) should start like this: This means that every sentence.
The study of how words combine to form grammatical sentences.
Long ago, and, oh, so far away I fell in love with you before the second show.
Introduction and Jurafsky Model Resource: A Probabilistic Model of Lexical and Syntactic Access and Disambiguation, Jurafsky 1996.
RELATIVE PRONOUNS DEFINING RELATIVE CLAUSES NON-DEFINING RELATIVE CLAUSES.
Focus On Grammar Book 2, 5 th edition Lesson 11: Adjective Clauses!!!!
Syntax: The Sentence Patterns of Language Deny A. Kwary Airlangga University.
Context-Free Parsing. 2/37 Basic issues Top-down vs. bottom-up Handling ambiguity –Lexical ambiguity –Structural ambiguity Breadth first vs. depth first.
Psy1302 Psychology of Language Lecture 10 Ambiguity Resolution Sentence Processing I.
PSY 369: Psycholinguistics Language Comprehension: Sentence comprehension.
Amirkabir University of Technology Computer Engineering Faculty AILAB Efficient Parsing Ahmad Abdollahzadeh Barfouroush Aban 1381 Natural Language Processing.
Matakuliah: G0922/Introduction to Linguistics Tahun: 2008 Session 11 Syntax 2.
1/13 Parsing III Probabilistic Parsing and Conclusions.
1/17 Probabilistic Parsing … and some other approaches.
Syntax: The analysis of sentence structure
PSY 369: Psycholinguistics Language Comprehension: Sentence comprehension.
You probably find many new words when you do your English activities. Don’t panic. Use a strategy to learn the new words. Avoid writing all the unknown.
Syntax: The analysis of sentence structure
SI485i : NLP Set 9 Advanced PCFGs Some slides from Chris Manning.
The Language Instinct Talking Heads.
Intro to NLP - J. Eisner1 Earley’s Algorithm (1970) Nice combo of our parsing ideas so far:  no restrictions on the form of the grammar:  A.
CS : Speech, Natural Language Processing and the Web/Topics in Artificial Intelligence Pushpak Bhattacharyya CSE Dept., IIT Bombay Lecture 12: Deeper.
1 Statistical Parsing Chapter 14 October 2012 Lecture #9.
Constituents Introduction to syntax. Review 打進季後賽 找到適當的理由 發現任何外傷 他送媽媽一束花 建議考生把握最後三周的時間 屬於那個國度.
1 Natural Language Processing Lecture 11 Efficient Parsing Reading: James Allen NLU (Chapter 6)
Natural Language Processing Lecture 6 : Revision.
Writing Sentences Gooder And to be most clearerest.
Verbs and prepositions Reported speech. Match the verbs with the prepositions and phrases: Respond Listen Deal Believe Delegate Communicate Invest in.
Language comprehension. understanding speech 1.differentiating speech sounds from other noises 2.recognizing words 3.activating their syntactic and semantic.
LINGUA INGLESE 1 modulo A/B Introduction to English Linguistics prof. Hugo Bowles Lesson 8 Syntax 1 1.
Business Concepts Expectations For: Classroom/ Students/ Teacher Fall 2014.
PS: Introduction to Psycholinguistics Winter Term 2005/06 Instructor: Daniel Wiechmann Office hours: Mon 2-3 pm Phone:
Today Phrase structure rules, trees Constituents Recursion Conjunction
.…… trust ……?. It’s finally almost here, mis quince años. So many things are going to change. I really hope everything turns out all right. I can already.
Subordinating Conjunctions
Review of basic concepts.  The knowledge of sentences and their structure.  Syntactic rules include: ◦ The grammaticality of sentences ◦ Word order.
Rules, Movement, Ambiguity
The man bites the dog man bites the dog bites the dog the dog dog Parse Tree NP A N the man bites the dog V N NP S VP A 1. Sentence  noun-phrase verb-phrase.
Adjectives. An Adjective is a word used to modify a noun or pronoun They tell what kind, which one, or how many. – Examples: old trophies, this classroom,
Exercises: Where is the bank ? -----I don’t know where the bank is What is he doing ? I don’t know what he is doing Why.
Syntax II “I really do not know that anything has ever been more exciting than diagramming sentences.” --Gertrude Stein.
STRESS AND INTONATION TEACHERS C1. Content and function words  Nouns : John, room, answer  Adjectives : happy, new, large, gray  Verbs : search, grow,
Run Why do you run, why do you hide, Oh don’t you know I just, I just want to be with you, To be with you.
SYNTAX.
CS460/IT632 Natural Language Processing/Language Technology for the Web Lecture 13 (17/02/06) Prof. Pushpak Bhattacharyya IIT Bombay Top-Down Bottom-Up.
Marking Homework 1. A very inexperienced journalist wrote the article about the accident. 2. The detective will give you a report about his investigation.
Sentence Types. Simple Sentence  A simple sentence has the most basic elements that make it a sentence: a subject, a verb, and a completed thought. 
…he said..  The students said: “Thank you for letting us in”  They all thanked the teacher for letting them in.
What did they say? Reported statements. Present Simple- Past Simple Present Continuous- Past Continuous Present Perfect- Past Perfect Past Simple- Past.
Noun Clauses A clause is a structure that has a subject and a verb. There are two kinds of clauses: independent and dependent. An independent clause is.
CS : Language Technology for the Web/Natural Language Processing Pushpak Bhattacharyya CSE Dept., IIT Bombay Parsing Algos.
Week 3. Clauses and Trees English Syntax. Trees and constituency A sentence has a hierarchical structure Constituents can have constituents of their own.
YESTERDAY.
Natural Language Processing
CS : Speech, NLP and the Web/Topics in AI
Subordinating Conjunctions
CSCI 5832 Natural Language Processing
Earley’s Algorithm (1970) Nice combo of our parsing ideas so far:
What did they say? Reported statements.
When we want to give extra information about the subject or the object of a sentence we need to add extra clauses. These clauses are often added on, or.
I CAN DECIDE…IT’S UP TO ME
Transitions in writing
YESTERDAY.
Presentation transcript:

Incremental parsing and ambiguous sentences - each incoming materials are attached immediately - a single analysis is pursed - given the uncertainty, how do we decide which structure to choose? Minimal attachment strategy Late closure strategy

The Temporary ambiguity

Theteacher Temporary ambiguity

Theteachersent Temporary ambiguity

Theteachersentthe Temporary ambiguity

Theteachersenttheflowers Temporary ambiguity

Theteachersentthewas pleasedflowers Temporary ambiguity

S NP DetN The teachersent All right, I have a verb. That’s easy. Let’s attach that under the main VP. Temporary ambiguity

S NP DetN VP The teachersent V Good. Then, I have to wait for some nouns after the verb. Come on, show me an NP now. Temporary ambiguity

S NP DetN VP NP DetN The teacherthe flowerssent V A-ha. You know, I’m good. So now what, maybe the recipient of the flowers? That would be a PP (yeah, to NP). Temporary ambiguity

S NP DetN VP NP DetN The teacherthe flowerssent V was pleased What????? That’s not possible. I already have a VP. How can I have another VP here? Somebody help me! Temporary ambiguity

S NP DetN The teachersent Don’t panic. Let’s go back. Maybe I did something wrong right here. Temporary ambiguity

S NPVP NPS DetNCompS NPVP NP DetN The teacher i who i eiei wasthe flowerswas pleasedsent Because I remember that there is another possibility. Suppose I have this structure. Then, Temporary ambiguity

S NPVP NPS DetNCompS NPVP NP DetN The teacher i the flowerswas pleasedsent who was I know that who was can be dropped without causing any meaning difference. Let’s see what happens. Temporary ambiguity

S NPVP NP DetN VP NP DetN The teacherthe flowerswas pleasedsent Now, it looks good. But wait, who would have thought of this structure in the first place? I wouldn’t! Temporary ambiguity

Theteachergiventhewas pleasedflowers Unambiguous sentence

S NPVP NPS DetNCompS NPVP NP DetN The teacher i who i eiei wasthe flowerswas pleasedgiven From the outset, it is clear that this is the only possible structure. Unambiguous sentence

S NPVP NPS DetNCompS NPVP NP DetN The teacher i the flowerswas pleasedgiven who was Again, who was can be dropped without causing any meaning difference. Unambiguous sentence

S NPVP NP DetN VP NP DetN The teacherthe flowerswas pleasedgiven The difference is that “given” is not ambiguous. Unambiguous sentence

S NPVP DetNVNP NDet The cop saw the spy So far so good! Global ambiguity

S NPVP DetNVNP NDet The cop saw the spy PP NPP with the binoculars Global ambiguity Two possible attachment sites -> how do I decide?

S NPVP DetNVNP NDet The cop saw the spy PP NPP with the binoculars Global ambiguity VP attachment  Minimal attachment strategy

S NPVP DetN V NP NDet The cop saw the spy NPPP NPP with the binoculars Global ambiguity NP attachment  more syntactic nodes

last night AdvP Global ambiguity Again two possible attachment sites  which one do I like better? S NPVP DetNVS N The reporter said the plane NP crashed VP V

Global ambiguity Late closure strategy  attach the incoming materials under the node currently being built S NPVP DetNVS N The reporter said the plane NPVP crashed V last night AdvP

last night AdvP Global ambiguity The attachment to the higher VP is not preferred S NPVP DetNVS N The reporter said the plane NP crashed VP V

While

Mary

WhileMarywas mending

WhileMarywas mendingthe sock

WhileMarywas mendingthe sockfell

WhileMarywas mendingthe sockfelloff

WhileMarywas mendingthe sockfelloffher lap.

Kensatsukan-ga prosecutor-nom Japanes e

Kensatsukan-ga prosecutor-nom kousakuin-o operative-acc Japanes e

Kensatsukan-ga prosecutor-nom kousakuin-o operative-acc goumonshita tortured Japanes e

Kensatsukan-ga prosecutor-nom kousakuin-o operative-acc goumonshita tortured jaajarisuto-o journalist-acc Japanes e

Kensatsukan-ga prosecutor-nom kousakuin-o operative-acc goumonshita tortured jaajarisuto-o journalist-acc Kisoshita indicted The prosecutor indicted the journalist who tortured the operative. Japanes e

S NPVP NPV operative tortured Yeah, this should be it. prosecutor Japanes e

S NPVP RC operative tortured Although this is possible, it looks much more complicated… prosecutor S N NPVP NPV eiei Japanes e

S NPVP NPV operative torturedprosecutor Japanes e journalist Oops, my previous analysis should be revised to …

S NPVP RC V operative tortured this one, the more complicated one. Otherwise, “journalist” has no place to attach itself to. prosecutor S N NPVP NPV eiei journalist i indicted Japanes e