1 Natural Language Processing Computational Discourse

1

Natural Language Processing

Computational Discourse

2

Outline

• Reference– Kinds of reference phenomena– Constraints on co-reference– Preferences for co-reference– The Lappin-Leass algorithm for

coreference

Note

• Reference resolution and discourse processing are very difficult

• There are no fully successful approaches• But, they are needed.• Today, the first goal is for you to be

aware of how you resolve pronouns (as a human). You need to understand the task before you could develop a method

• Then, we will look at a concrete algorithm for pronoun resolution

• Finally, we’ll look at discourse relations

3

4

Part I: Reference Resolution

• John went to Bill’s car dealership to check out an Acura Integra. He looked at it for half an hour

• I’d like to get from Boston to San Francisco, on either December 5th or December 6th. It’s ok if it stops in another city along they way

5

Some terminology

• John went to Bill’s car dealership to check out an Acura Integra. He looked at it for half an hour

• Reference: process by which speakers use words John and he to denote a particular person– Referring expression: John, he– Referent: the actual entity (but as a shorthand we

might call “John” the referent).– John and he “corefer”– Antecedent: John– Anaphor: he

6

Discourse Model

• Model of the entities the discourse is about

• A referent is first evoked into the model. Then later it is accessed from the model

John HeCorefer

Evoke Access

7

Many types of reference

• (after Webber 91)• According to John, Bob bought Sue an

Integra, and Sue bought Fred a Legend– But that turned out to be a lie (a speech act)– But that was false (proposition)– That struck me as a funny way to describe

the situation (manner of description)– That caused Sue to become rather poor

(event)– That caused them both to become rather

poor (combination of several events)

8

Reference Phenomena

• Indefinite noun phrases: generally new– I saw an Acura Integra today– Some Acura Integras were being unloaded…– I am going to the dealership to buy an Acura Integra

today. (specific/non-specific)• I hope they still have it• I hope they have a car I like

• Definite noun phrases: identifiable to hearer because– Mentioned: I saw an Acura Integra today. The

Integra was white– Identifiable from beliefs: The Indianapolis 500– Inherently unique: The fastest car in …

9

Reference Phenomena: Pronouns

• I saw an Acura Integra today. It was white• Compared to definite noun phrases,

pronouns require more referent salience.– John went to Bob’s party, and parked next to a

beautiful Acura Integra– He got out and talked to Bob, the owner, for more

than an hour.– Bob told him that he recently got engaged and

that they are moving into a new home on Main Street.

– ??He also said that he bought it yesterday.– He also said that he bought the Acura yesterday

10

Salience Via Structural Recency

• E: So, you have the engine assembly finished. Now attach the rope. By the way, did you buy the gas can today?

• A: Yes• E: Did it cost much?• A: No• E: Good. Ok, have you got it attached

yet?• The middle part is a subdialog; once it is

over, we resume with the discourse model of the first two sentences. “By the way” signals the beginning of the subdialog and “Ok” signals the end of it.

11

More on Pronouns

• Cataphora: pronoun appears before referent:– Before he bought it, John checked over

the Integra very carefully.

12

Inferrables

• I almost bought an Acura Integra today, but the engine seemed noisy..

• Mix the flour, butter, and water.– Kneed the dough until smooth and shiny– Spread the paste over the blueberries– Stir the batter until all lumps are gone.

13

Inferrables

• I almost bought an Acura Integra today, but the engine seemed noisy. This works, even if the engine wasn’t mentioned before.

• Mix the flour, butter, and water. Same for these.– Kneed the dough until smooth and shiny– Spread the paste over the blueberries– Stir the batter until all lumps are gone.

14

Discontinuous sets

• John has an Acura and Mary has a Suburu. They drive them all the time.

15

Generics

• I saw no less than 6 Acura Integras today. They are the coolest cars.

16

Generics

• I saw no less than 6 Acura Integras today. They are the coolest cars.

• “They” refers to the general class, not to the 6 specific cars

17

Pronominal Reference Resolution

• Given a pronoun, find the referent (either in text or as a entity in the world)

• We will approach this today in 3 steps– Hard constraints on reference– Soft constraints on reference– An algorithm that uses these constraints

• But first, let’s think about what influences pronominal reference resolution

18

What influences pronoun resolution?

• Syntax• Semantics/world knowledge

19

Syntax matters

• John kicked Bill. Mary told him to go home.

• Bill was kicked by John. Mary told him to go home.

• John kicked Bill. Mary punched him.

20

Why syntax matters

• John kicked Bill. Mary told him to go home. John; subject preference.



21

Why syntax matters


• Bill was kicked by John. Mary told him to go home. Bill; subject preference. This is due to syntax, not semantics, since it is the same event.


22

Why syntax matters



• John kicked Bill. Mary punched him. Bill; parallel roles (both are direct objects); that can trump the subject preference

23

Why syntax matters




Grammatical role hierarchy

24

Why syntax matters




Grammatical role

parallelism

25

Why semantics matters

The city council denied the demonstrators a permit because they {feared|advocated} violence.

26



27



28

Why knowledge matters

• John hit Bill. He was severely injured.

29

Why knowledge matters

• John hit Bill. He was severely injured.

• Even though John is the subject, our knowledge tells us “He” is probably Bill, overriding the syntactic preference

30

Margaret Thatcher admires Hillary Clinton, and George

W. Bush absolutely worships her.

Why Knowledge Matters

31

Margaret Thatcher admires Hillary Clinton, and George W. Bush

absolutely worships her.

Remember: George W. Bush is positive toward Thatcher and negative toward H. Clinton; so we reject

“her” being H. Clinton, even though syntactic parallelism suggests it is Clinton. (Or, you consider

sarcasm, that the speaker is wrong, etc.)

Why Knowledge Matters

Pronoun Resolution: State of the Art

• Not very good• There are no successful methods that

take care of all the different cases we’ve seen in these notes (e.g., those that require knowledge, detecting subdialogs, detecting sarcasm, etc)

• There are approaches that exploit hard and soft syntactic and semantic constraints

• In general, we need reference resolution to succeed to get to the “next level” of NLP systems. Hopefully, there will be a breakthrough …

32

33

Hard constraints on coreference

• Number agreement– John has an Acura. It is red.

• Person and case agreement– *John and Mary have Acuras. We love them

(where We=John and Mary)

• Gender agreement– John has an Acura. He/it/she is attractive.

• Syntactic constraints– John bought himself a new Acura (himself=John)– John bought him a new Acura (him = not John)

34

Pronoun Interpretation Preferences

• Selectional Restrictions– John parked his Acura in the garage. He

had driven it around for hours.

• Recency– John has an Integra. Bill has a Legend.

Mary likes to drive it.

35

Pronoun Interpretation Preferences

• Grammatical Role: Subject preference– John went to the Acura dealership with

Bill. He bought an Integra.– Bill went to the Acura dealership with

John. He bought an Integra– (?) John and Bill went to the Acura

dealership. He bought an Integra

36

Repeated Mention preference

• John needed a car to get to his new job. He decided that he wanted something sporty. Bill went to the Acura dealership with him. He bought an Integra.

• “He” is John, even though Bill was more recently mentioned. But, John is more “salient” (in focus) because he was mentioned 3 times compared to 1 time for Bill.

37

Parallelism Preference

• Mary went with Sue to the Acura dealership. Sally went with her to the Mazda dealership. (even though Mary is the subject)

• Mary went with Sue to the Acura dealership. Sally told her not to buy anything. (no parallelism, so subject preferred)

38

Verb Semantics Preferences

• John telephoned Bill. He lost the pamphlet on Acuras.

• John criticized Bill. He lost the pamphlet on Acuras.

• “Implicit causality”– Implicit cause of criticizing is object.– Implicit cause of telephoning is subject.– (To understand this, try to be aware of

how you are interpreting the sentences.)

39

Verb Preferences

• John seized the Acura pamphlet from Bill. He loves reading about cars.

• John passed the Acura pamphlet to Bill. He loves reading about cars.

40

Verb Preferences

• John seized the Acura pamphlet from Bill. He loves reading about cars.

• After the first sentence, John is the focus – you picture him seizing the pamphlet and bringing it toward himself.

• John passed the Acura pamphlet to Bill. He loves reading about cars.

• After the first sentence, Bill is the focus; he just received the pamphlet

• Preference for goal over source semantic role

41

Pronoun Resolution Algorithm

• Lappin and Leass (1994): Given he/she/it, assign antecedent.

• Implements only recency and syntactic preferences

• Two steps– Discourse model update

• When a new noun phrase is encountered, add a representation to discourse model with a salience value

• Modify saliences.

– Pronoun resolution• Choose the most salient antecedent

42

Salience Factors and Weights

• From Lappin and Leass

Sentence recency 100

Subject emphasis 80

Existential emphasis 70

Accusative (direct object) emphasis

50

Ind. Obj and oblique emphasis 40

Non-adverbial emphasis 50

Head noun emphasis 80

43

Recency

• Weights are cut in half after each sentence is processed

• A sentence recency weight is added (100 for new sentences)

44

Lappin and Leass (cont)

• Grammatical role preference– Subject > existential predicate nominal > object >

indirect object > demarcated adverbial PP

• Examples– An Acura Integra is parked in the lot (subject)– There is an Acura Integra parked in the lot (ex. pred

nominal)– John parked an Acura Integra in the lot (object)– John gave his Acura Integra a bath (indirect obj)– In his Acura Integra, John showed Susan his new CD

player (demarcated adverbial PP)

• Head noun emphasis factor gives above 80 points, but followed embedded NP nothing:– The owner’s manual for an Acura Integra is on John’s

desk

45

Lappin and Leass (cont)

• Grammatical role preference– Subject > existential predicate nominal > object >

indirect object > demarcated adverbial PP

• Examples– In his Acura Integra, John showed Susan his new CD

player (demarcated adverbial PP) – a noun inside a PP which is “demarcated” – separated from the main sentence with a comma. Note that any noun that is NOT in one of these gets 50 points.

• Head noun emphasis factor gives 80 points to a noun that is the head of its NP. The following does not get these points:– The owner’s manual for an Acura Integra is on John’s

desk

46

Lappin and Leass Algorithm

• Collect the potential referents (up to 4 sentences back)

• Remove potential referents that do not agree in number or gender with the pronoun

• Remove potential references that do not pass syntactic coreference constraints

• Compute total salience value of referent from all factors, including, if applicable, role parallelism (+35) or cataphora (-175).

• Select referent with highest salience value. In case of tie, select closest.

• (We’ll fill in specifics of what happens when via example)

47

Example

• John saw a beautiful Acura Integra at the dealership. He showed it to Bob. He bought it.

rec Subj Exist

Obj Ind-obj

Non-adv

Head N

Total

John 100 80 50 80 310

Integra 100 50 50 80 280

dealership

100 50 80 230

Sentence 1 (no pronoun to resolve):

48

After sentence 1

• Cut all values in half

Referent Phrases Value

John {John} 155

Integra {a beautiful Acura Integra}

140

dealership

{the dealership} 115

49

He showed it to Bob

• He specifies male gender• So Step 2 reduces set of referents to only

John.• Now update discourse model for the

referent:• He in current sentence (recency=100),

subject position (=80), not adverbial (=50) not embedded (=80), so add 310:Referent Phrases Value

John {John, he1} 155+310

Integra {a beautiful Acura Integra}

140

dealership

{the dealership 115

50

He showed it to Bob

• Can be Integra or dealership.• Need to add weights:

– Parallelism: it + Integra are objects (dealership is not), so +35 for integra

– Integra 175 to dealership 115, so pick Integra

• Update discourse model: it is nonembedded object, gets 100+50+50+80=280 added to it (in current sentence, direct object, not in a PP, is the head noun). It had 140, so now its total is 420.

51

He showed it to Bob (after “it”, before “Bob”)


John {John, he1} 465

Integra {a beautiful Acura Integra, it1}

420

dealership


52

He showed it to Bob

• Bob is new referent, is oblique argument, weight is 100+40+50+80=270 (on the exam, if something is an oblique argument, I’ll tell you)Referent Phrases Valu

e

John {John, he1} 465


420

Bob {Bob} 270

dealership


53

He bought it

• Drop weights in half:


John {John, he1} 232.5


210

Bob {Bob} 135

dealership

{the dealership} 57.5He2 will be resolved to John: John has 232.5 + 35for parallelism (he showed it to bob; he bought it)

54

He bought it

• Resolved to Integra – 210 + 35 for parallelism (He showed it to bob; he bought it):


John {John, he1,he2} 232.5+100+80+50+80


210

Bob {Bob} 135

dealership

{the dealership} 57.5

55

He bought it (after “it” before going on to the next sentence)

• Drop weights in half:


John {John, he1,he2} 542.5

Integra {a beautiful Acura Integra, it1,it2}

210+100 + 50 + 50 + 80

Bob {Bob} 135

dealership

{the dealership} 57.5

56

Reference Resolution

• Many other algorithms and other constraints– Centering theory: constraints which focus

on discourse state, and focus. – Hobbs: ref. resolution as by-product of

general reasoning– Mitkov et al. (e.g.) Machine learning

Documents

1 Natural Language Processing Computational Discourse