Date post: | 05-Jan-2016 |
Category: |
Documents |
Upload: | victor-bradley |
View: | 215 times |
Download: | 3 times |
Coherence and Coreference
Introduction to Discourse and Dialogue
CS 359
October 2, 2001
Publicly Available Telephone Demos• Nuance http://www.nuance.com/demo/index.html
– Banking: 1-650-847-7438– Travel Planning: 1-650-847-7427– Stock Quotes: 1-650-847-7423
• SpeechWorks http://www.speechworks.com/demos/demos.htm– Banking: 1-888-729-3366– Stock Trading: 1-800-786-2571
• MIT Spoken Language Systems Laboratory http://www.sls.lcs.mit.edu/sls/whatwedo/applications.html– Travel Plans (Pegasus): 1-877-648-8255– Weather (Jupiter): 1-888-573-8255
• IBM http://www.software.ibm.com/speech/overview/business/demo.html– Mutual Funds, Name Dialing: 1-877-VIA-VOICE
From Caroenter and Chu-Carroll, Tutorial on Spoken Dialogue Systems, ACL ‘99
Discussion questions
• What to say/how to say it distinction: Part of determining “how to say it” necessarily depends on “reading” the hearer accurately. To what extent could a computer system gauge the myriad factors - expression, body language, gesture, past utterances - to “read” the hearer? Is it a question of understanding, programming or processing?
Discussion questions
• How is a set of texts chosen? What makes a text good for this type of analysis? Why recipes?
• How could a system cope with anaphora when there is insufficient information to resolve it at the time of utterance?
• How well do systems really do at resolving extended chains of reference?
• How would these systems deal with the more complex hierarchical, embedded discourse structures that we see in the real world?
Agenda
• Coherence: Holding discourse together– Coherence types and relations
• Reference resolution– Syntactic & semantic constraints– Syntactic preferences– A first resolution algorithm
Coherence: Holding Discourse Together
• Cohesion: – Necessary to make discourse a semantic unit– All utterances linked to some preceding utterance– Expresses continuity
– Key: Enables hearers to interpret missing elements, through textual and environmental context links
Cohesive Ties (Halliday & Hasan, 1972)
• “Reference”: e.g. “he”,”she”,”it”,”that”– Relate utterances by referring to same entities
• “Substitution”/”Ellipsis”:e.g. Jack fell. Jill did too.– Relate utterances by repeated partial structure w/contrast
• “Lexical Cohesion”: e.g. fell, fall, fall…,trip..– Relate utterances by repeated/related words
• “Conjunction”: e.g. and, or, then– Relate continuous text by logical, semantic, interpersonal relations.
Interpretation of 2nd utterance depands on first
Reference Resolution
• Match referring expressions to referents
• Syntactic & semantic constraints
• Syntactic & semantic preferences
• A 1st resolution algorithm
Reference (terminology)
• Referring expression: (refexp)– Linguistic form that picks out entity in some model– That entity is the “referent”
• When introduces entity, “evokes” it
• Set up later reference, “antecedent”
– 2 refexps with same referent “co-refer”
• Anaphor:– Abbreviated linguistic form interpreted in context– Refers to previously introduced item (“accesses”)
Referring Expressions• Indefinite noun phrases (NPs): e.g. “a cat”
– Introduces new item to discourse context
• Definite NPs: e.g. “the cat”– Refers to item identifiable by hearer in context
• By verbal, pointing, or environment availability
• Pronouns: e.g. “he”,”she”, “it”– Refers to item, must be “salient”
• Demonstratives: e.g. “this”, “that”– Refers to item, sense of distance (literal/figurative)
• One-anaphora: “one” – One of a set, possibly generic
Syntactic Constraints
• Agreement:– Number: Singular/Plural
– Person: 1st: I,we; 2nd: you; 3rd: he, she, it, they
– Case: we/us; he/him; they/them…
– Gender: he vs she vs it
Syntactic & Semantic Constraints
• Binding constraints:– Reflexive (x-self): corefers with subject of clause– Pronoun/Def. NP: can’t corefer with subject of clause
• “Selectional restrictions”:– “animate”: The cows eat grass.– “human”: The author wrote the book.– More general: drive: John drives a car….
Syntactic & Semantic Preferences
• Recency: Closer entities are more salient
• Grammatical role: Saliency hierarchy of roles– e.g. Subj > Object > I. Obj. > Oblique > AdvP
• Repeated reference: Pronouns more salient
• Parallelism: Prefer entity in same role
• Verb roles: “implicit causality”, thematic role match,...
Reference Resolution Approaches
• Common features– “Discourse Model”
• Referents evoked in discourse, available for reference
• Structure indicating relative salience
– Syntactic & Semantic Constraints– Syntactic & Semantic Preferences
• Differences:– Which constraints/preferences? How combine?
Rank?
A Resolution Algorithm
• Discourse model update:– Evoked entities:
• Equivalence classes: Coreferent referring expressions
– Salience value update:• Weighted sum of salience values:
– Based on syntactic preferences
• Pronoun resolution:– Exclude referents that violate syntactic constraints– Select referent with highest salience value
Salience Factors (Lappin & Leass 1994)
• Weights empirically derived from corpus• Recency: 100• Subject: 80• Existential: 70• Object: 50• Indirect Object/Oblique: 40• Non-adverb PP: 50• Head noun: 80• Parallelism: 35, Cataphora: -175
– Divide by 50% for each sentence distance
Example
• John saw a beautiful Acura Integra in the dealership.
• He showed it to Bob.
• He bought it.
Example
• John saw a beautiful Acura Integra in the dealership.
Referent Phrases ValueJohn {John} 310Integra {a beautiful Acura Integra} 280dealership {the dealership} 230
Example
• He showed it to Bob.
Referent Phrases ValueJohn {John, he1} 465Integra {a beautiful Acura Integra} 140dealership {the dealership} 115
Referent Phrases ValueJohn {John, he1} 465Integra {a beautiful Acura Integra, it1} 420dealership {the dealership} 115
Example
• He showed it to Bob.
Referent Phrases ValueJohn {John, he1} 465Integra {a beautiful Acura Integra, it1} 420Bob {Bob} 270dealership {the dealership} 115
Example
• He bought it.
Referent Phrases ValueJohn {John, he1} 232.5Integra {a beautiful Acura Integra, it1} 210Bob {Bob} 135dealership {the dealership} 57.5
Referent Phrases ValueJohn {John, he1, he2} 542.5Integra {a beautiful Acura Integra, it1, it2} 520Bob {Bob} 135dealership {the dealership} 57.5
Coherence & Coreference
• Cohesion: Establishes semantic unity of discourse– Necessary condition– Different types of cohesive forms and relations– Enables interpretation of referring expressions
• Reference resolution– Syntactic/Semantic Constraints/Preferences– Discourse, Task/Domain, World knowledge
• Structure and semantic constraints
Challenges
• Alternative approaches to reference resolution– Different constraints, rankings, combination
• Different types of referent– Speech acts, propositions, actions, events– “Inferrables” - e.g. car -> door, hood, trunk,..– Discontinuous sets– Generics– Time