Information Structure and its Use for Context-Sensitive

R

A VIE N

Information Structure and its Use for Context-Sensitive Information Presentation Ivana Kruijff-Korbayová [email protected] http://www.coli.uni-sb.de/˜korbay ACL/HCSNet Advanced Program in NLP University of Melbourne, 10-14 July 2006

I.Kruijff-Korbayová

IS & Presentation

S

SA

IS

S

UN

E R SIT A

IV

ACL/HCSNet NLP Program 2006

R

S

IS

S

UN

E R SIT A

SA

1

IV

A VIE N

Outline • • • • • • • •

Motivation Information Structure Partitioning Bird’s-Eye View on IS Theories and Terminologies Question test for IS IS Realization Means IS Semantics Meaning Differences due to IS IS Applications in Implemented Systems – Answers to Questions and Dialogue – MT, Text Generation and TTS – Multimodal Dialogue • Wrap Up I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

2

IV

A VIE N

Motivation


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

3

IV

A VIE N

Illustration (1) Sign in London underground: Dogs must be carried. (Halliday, 1970) a. Dogs must be carried. Interpretation: If you have a dog, you must carry it. b. Dogs must be carried. Interpretation: You must carry a dog. But: (2) Sign in a synagogue: Hats must be worn. a. Hats must be worn. Interpretation: If you have a hat, you must wear it. b. Hats must be worn. Interpretation: You must wear a hat. (Capitals denote main intonation center = nuclear stress in the sentence.) I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

4

IV

A VIE N

Illustration (3)

a. U: What is the status of the stove? S: The stove is switched on. b. U: Which device is switched on? S: The stove is switched on.

(4)

a. U: When does Jane have swimming on Tuesday? S: Jane has swimming at noon on Tuesday. S0: On Tuesday, Jane has swimming at noon. b. U: What does Jane have at noon on Tuesday? S: Jane has swimming at noon on Tuesday. S0: At noon on Tuesday, Jane has swimming.

Various aspects of surface form are involved! I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

5

IV

A VIE N

Illustration (5) Czech: a. Psi se mus´ı n´ est. Dogsnom refl must3pl carryinf b. Mus´ı se nést pes. Must3sg refl carryinf dognom (6) German: a. Hunde mussen getragen werden. Dogs must carried be b. Es mussen Hunde getragen werden. Itexplet must3pl dogsnom carriedpart beinf . Different languages may realize the same meaning differences differently! I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

6

IV

A VIE N

Observations • Surface form of utterances reflects their communicative goal w.r.t. the context – Variation of surface form aspects, such as intonation, word order and syntactic structure is not arbitrary – Not just what is said but also how is important (cf. Grice’s Maxim of Manner (Grice, 1975)) • Various aspects of surface form interact • Different languages may use different combinations of surface form aspects to realize the same variation ⇒ Information Structure as a unifying abstract notion at the level of utterance meaning I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

7

IV

A VIE N

Information Structure Partitioning


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

8

IV

A VIE N

Information Structure • IS concerns a division (partitioning) (Kruijff-Korbayová and Steedman, 2003):

of

an

utterance

meaning

– Theme the part which relates it to the purpose of the discourse and anchors the content to the context (i.e., what speaker and hearer are attending to); what the utterance is about, the topic that the speaker means to address; may also restrict the context to particular type(s) of situation(s) – Rheme the part which advances the discourse, i.e., adds or modifies some information (i.e., the informative part); what the speaker says about the Theme, i.e., the Rheme is semantically predicated over the Theme • IS is an inherent aspect of meaning —it is an important factor in establishing coherence with respect to the context in which a sentence is uttered. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

9

IV

A VIE N

Theme-Rheme Partitioning: Examples (7)

a.

Dogs must be carried. | {z } | {z } T heme

b. |

Rheme

Dogs must be carried. {z } Rheme

c.

Dogs must be carried. {z } | {z } | Rheme


T heme

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

10

IV

A VIE N

Theme-Rheme Partitioning: Examples (8)

a. U: What is the status of the stove? The stove is switched on. | {z }| {z } T heme

Rheme

b. U: Which device is switched on? S: The stove is switched on. | {z }| {z } Rheme


T heme

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

11

IV

A VIE N

IS Theories


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

12

IV

A VIE N

IS Theories Phenomena known at least since mid 19th century (France, Germany) Since then, essentially three classes of theories: 1. IS is described from the viewpoint of semantics, in a “degrammatized” way —mostly theories following up on Karttunen, Cresswell, von Stechow . . . 2. Only the realization of IS (syntax, intonation) is described —mostly Generative Grammar-based theories, following up on Chomsky, Bolinger . . . 3. Theories that stress the integration, into a single framework, of the “semantics” of IS and its realization —such as Sgall et al., Halliday, Vallduv´ı, Steedman ...


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

13

IV

A VIE N

The following diagram displays a view of the influences and terminological dependencies in theories of IS, and their links to theories of discourse structure and discourse semantics, as presented in (Steedman and Kruijff-Korbayová, 2001)


IS & Presentation


A

R

S

IS

S

UN

E R SIT

A VIE N

(Russell 1905)

Mathesius 1929 nucleus/focus known/unknown

(Strawson 1950, 1954) presupposition

Firbas 1964, 1966 theme/rheme context dependent/independent Halliday 1967 theme’/rheme’ given/new (orthogonal)

Bolinger 1965 theme/rheme, accent Chomsky 1965 Sgall 1967 topic/comment topic/focus, context bound/unbound Karttunen 1968

(Sacks, Schegloff & Jefferson 1974)

Chomsky 1970/Jackendoff 1970 presupposition/focus topic/comment (orthogonal)

Dahl 1969 topic/comment (Winograd, Woods) background/focus Kay 1975 (Halliday & Hasan 1976) given/new (Grimes 1975) Chafe, Clark, Gundel, Prince topic/comment given/new’ (orthogonal) (Polanyi and Scha 1983 ) (Brown 1983)

(Mann & Thompson 1987)

SA

14

IV

(Montague 1973)

(Cresswell, von Stechow Karttunen & Peters 1979 Kamp, Heim) presupposition/focus (structured meanings, (alternative set) DRT) Selkirk 1984

(Grosz & Sidner, Webber) (Pierrehumbert & Hirschberg, Grosz, Joshi & Weinstein) Steedman 1991 Vallduvi 1990 theme/rheme, link/tail/focus background/focus Vallduvi & Vilkuna 1998 theme/rheme, 0/kontrast

Krifka, Kratzer Rooth 1985 presupposition/narrow focus, wide focus .. Buring 1995 topic/focus C/Q alternatives set

Hajicova, Partee, & Sgall 1998 topic/focus, context bound/unbound

Hendriks 1999 link/tail/focus


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

15

IV

A VIE N

Aligning IS Terminologies Mathesius, Firbas, Daneˇs Sgall et al.

Halliday Chomsky, Jackendoff, Krifka Rooth Vallduv´ı Steedman


Theme

vs.

Rheme

Topic (CB) topic proper vs. contrastive topic communicative dynamism [Theme] Given vs. New

vs.

Focus (NB) focus proper

Presupposition

vs.

Focus

Ground Tail vs. Link (Kontrast) Theme Background vs. Focus

vs.

Focus

[Rheme] Given vs. New

Kontrast vs.

IS & Presentation

Rheme Background vs. Focus ACL/HCSNet NLP Program 2006

R

S

IS

S

UN

E R SIT A

SA

16

IV

A VIE N

Aligning IS Terminologies But, be aware of differences concerning: • Definitions of the IS categories and operational criteria • Level(s) at which IS distinctions are made, e.g., surface, deep, meaning . . . • Flexible vs. fixed syntactic constituents, and how do IS components correspond to them • Degree of recursivity of IS notions (if any) • IS in complex sentences


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

17

IV

A VIE N

Question test for IS


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

18

IV

A VIE N

Question Test for IS Question-answer pairs are commonly used to indicate or (operationally) test the context in which a particular IS is appropriate: the question determines the Theme; what is asked for is the Rheme; the answer “fills” the Rheme. (Sgall et al., 1986; Steedman, 2000; Kruijff-Korbayová and Steedman, 2003) (9) Q. What does John do? A. John writes novels . | {z } | {z } T heme

Rheme

(10) Q. Who writes novels? A. John writes novels . | {z } | {z } Rheme

T heme

Swapping the questions in the two examples results in incoherent Q-A pairs: the answers become infelicitous, because the IS partitioning then does not match the context set by the question.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

19

IV

A VIE N

Question Test for IS: “Focus Projection” The linguistic form of an utterance may be compatible with several different IS partitionings (as reflected by different questions) (11)

a. What does John write? John writes novels. | {z }| {z } T heme

Rheme

b. What does John do? John writes novels. {z } | {z } | T heme

”narrow focus”

”broad focus”

Rheme

c. What is going on? / What is the news? John writes novels. | {z }

”broad focus”

Rheme


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

20

IV

A VIE N

(12) John writes novels. # Who writes novels? # What does John do with/about novels? (13) Who writes novels? John writes novels. | {z } | {z } Rheme

”narrow focus”

T heme

(14) What does John do with novels? John writes novels. | {z } | {z } | {z } T heme


Rheme

”narrow focus”

T heme

IS & Presentation


R

S

IS

S

UN

E R SIT A

21

SA

cont’d

IV

A VIE N

Question Test for IS: “Focus Projection” (15) John flew from London to Paris. (16) a. What happened? b. What did John do? c. Where did John fly? d. To which place did John fly from London? (15) is good as an answer to (16a-d). It has several possible IS partitionings.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

22

IV

A VIE N

(17) is only a matching answer to (18). (17) John flew to Paris from London. (18) Where did John fly from London? (19) is only a matching answer to (20). (19) John flew from London to Paris. (20) From which place did John fly to Paris? Neither of the above is a good answer to (21). (21) Who flew from London to Paris? (22) John flew from London to Paris.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

23

IV

A VIE N

Two Dimensions of IS Partitioning


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

24

IV

A VIE N

Two Dimensions of IS Partitioning IS concerns two aspects of a division (partitioning) of an utterance meaning (Kruijff-Korbayová and Steedman, 2003): • Theme-Rheme partitioning reflects the dichotomy between what an utterance is about (how its message is anchored to the context) and how it advances the discourse • Background-Focus partitioning reflects an abstract notion of contrast between alternatives available in the discourse context, against which the actual utterance is cast; (Steedman, 2000): B/F partitioning within Theme and/or Rheme reflects alternative Theme(s) and/or Rheme(s) in the context. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

25

IV

A VIE N

IS Partitioning: Background-Focus in Rheme (23) I know the british author writes history books. But what does he read? a. The british author reads comic books. | {z } | {z } T heme } | F ocus{z Rheme

b. |

The british author reads history papers . {z } | {z } T heme | {zF ocus } Rheme

c. |

The british author reads detective stories . {z } | {z } | {z } F ocus F ocus T heme {z } | Rheme


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

26

IV

A VIE N

(24) I know the american author reads comics. But who reads history books? a. The american actor reads history books. | {z } | {z } T heme {z F ocus } | Rheme

b.

The british author reads history books. | {z } {z } | F ocus T heme | {z } Rheme

c.

The british actor reads history books. | {z } | {z } | {z } F ocus F ocus T heme {z } | Rheme


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

27

IV

A VIE N

IS Partitioning: Background-Focus in Theme (25) I know the american author reads history books. But what does the british author read? The british author reads comics . | {z } | {z } F ocus | {z } | F ocus {z } T heme

Rheme

(26) I know the american author reads history books. But what does the american actor read? The american actor reads comics . | {z } | {z } | {z F ocus } | F ocus {z } T heme


Rheme

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

28

IV

A VIE N

IS Partitioning: Background-Focus in Theme and Rheme (27) I know the german author reads history novels. But what does the german actor read? detective novels. The german actor reads | | {z } {z } F ocus {zF ocus }| {z } | T heme

Rheme

(28) I know the british author writes history books. But what does he read? The british author reads comic books. | {z } | {z } F ocus {z } | F ocus{z } | T heme


Rheme

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

29

IV

A VIE N

IS Realization


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

30

IV

A VIE N

IS Realization Means • Various means that can be used to realize IS (i.e., IS influences various aspects of linguistic form) – intonation (prosody); e.g., the predominant means in (not only) English – (word) ordering; e.g. the predominant means in Czech and other Slavic languages, also to some extent in German (particularly in the “Mittelfeld”) – morphological/grammatical marking; e.g., particles ‘wa’ and ‘ga’ in Japanese – syntactic constructions, e.g. it-cleft, wh-cleft, passivization, etc. – ellipsis • The means can be used also in combination • Different languages employ and combine the means differently, depending on their typological characteristics (Vallduv´ı and Engdahl, 1996) I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

31

IV

A VIE N

IS Realization Means: Intonation (Steedman, 2000) for English; similarly (Uhmann, 1991; Fery, 1993) for German: • Theme/Rheme partitioning – – – –

Determines overall intonation pattern Theme and Rheme as one intonation phrase each (boundary between) Theme-accents: L+H*, L*+H (prototypical Theme-tune: L+H*LH%) Rheme-accents: H*, L*, H*+L, H+L* (prototypical Rheme-tune: H*LL%)

• Background/Focus partitioning – Determines placement of pitch accents on particular words – Focus: marked by pitch accent – Background: unmarked by pitch accent I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

32

IV

A VIE N

Example from (Kruijff-Korbayová et al., 2003) (29)U: S:

What devices are there in the house? There is a stove in the kitchen , H* H* | {z } | {z } F ocus F ocus | {z } Rheme

a radio in the kitchen and a radio in the bathroom . H* H*LL% | {z } | {z } F ocus F ocus | {z } Rheme


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

33

IV

A VIE N

U: What is the status of the kitchen devices? S: The stove in the kitchen is on . L+H* L% H*LL% | {z } | {z } | {z } F ocus F ocus Background {z } | {z }| Rheme

T heme

The radio L+H* | {z } F ocus

|

in the kitchen is off . H*LL% L% | {z } | {z } F ocus {z } {z Background } | Rheme

T heme


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

34

IV

A VIE N

IS Realization Means: Word Order “Normal” (default) order: Theme before Rheme (and Background before Focus) (30) What does John write? |

John writes novels. {z }| {z } T heme

Rheme

“Subjective ordering” (Firbas, 1971; Firbas, 1992): Rheme before Theme (31) What does John write? |

Novels John writes. {z } | {z } Rheme

T heme

—typically motivated by discourse context But, any IS-sensitive ordering must respect syntactic constraints!


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

35

IV

A VIE N

Theme before Rheme: German examples from an MP3-WOZ dialogue corpus (32) MP3: Was soll mit den Liedern gemacht werden? What should with the songs done become What should be done with the songs? U: Mit den Liedern soll eine Playlist erstellt werden. With the songs should a playlist created become From the songs should be created a playlist. (33) U: Bitte suche Titel von Madonna. Please search tracks from Madonna Please find tracks by Madonna. MP3: Einen Moment ... Von Madonna haben wir 1711 Treffer. A moment From Madonna have we 1711 hits Just a moment ... From Madonna we have 1711 hits. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

36

IV

A VIE N

IS Realization Means: Syntax Syntactic constructions that change order or provide “bracketing”: • • • • • • • •

fronting (so-called topicalization) left dislocation right dislocation cleft pseudo-cleft passivization dative shift there-insertion

Differences across languages! Differences in contextual appropriateness. I.Kruijff-Korbayová

IS & Presentation

(Prince, 1978) ACL/HCSNet NLP Program 2006

R

S

IS

S

UN

E R SIT A

SA

37

IV

A VIE N

Fronting and dislocation (of Theme): (34)

a. Novels, John writes. b. Novels, John writes them. c. John writes them, novels.

Cleft: It is Rheme (that/who) Theme (35)

a. What does John hate? It is comics John hates. b. Who hates comics? It is John who hates comics.

Pseudo-cleft: Who/What Theme is/are Rheme (36)

a. What does John hate? What John hates are comics. b. Who hates comics? Who hates comics is John.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

38

IV

A VIE N

Passivization: allows opposite ordering than corresponding active sentence (37) Who hates comics? a. Comics are hated by John. b. John hates comics. Dative shift: (38)

a. John gave a book to Mary. b. John gave Mary a book.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

39

IV

A VIE N

There-insertion: gets Rheme-subject away from the beginning of the sentence (39) What is in the garden? a. There is a troll in the garden. b. A troll is in the garden. (40) Example heard on German radio: Es spielt die Tschechische Philharmonie. There plays the Czech Philharmonic Now we will be playing the Czech Philharmonic. dirigiert Hilary Griffiths. Es There directs Hilary Griffiths It is directed by Hilary Griffiths. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

40

IV

A VIE N

IS Realization Means: Ellipsis What is “known” (available, retrievable) in context can be left out: (41) (42)

A: B: A: B:

What does John hate? Comics. Who hates comics? John.

Ellipsis example from the Map Task corpus (43) G: where are you in relation to the top of the page just now? F: Uh, about four inches. G: Four inches? F: Yeah. G: Where are you from the left-hand side? F: About two. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

41

IV

A VIE N

IS Semantics


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

42

IV

A VIE N

IS-Sensitive Interpretation • IS semantics using structured meanings (von Stechow, 1990; Jackendoff, 1990; Krifka, 1992; Krifka, 1993) • Semantics of (Vallduv´ı, 1992)

information

packaging

using

file-change

functions

• Semantics of questions in terms of answer-alternative sets (Hamblin, 1973) • Focus semantics in terms of focus-alternative set (Rooth, 1992) • Semantics of focus-marked topics in terms of question-alternative sets (B¨ uring, 1997; B¨ uring, 1999) • Semantics of two-dimensional IS-partitioning in terms of Rheme-alternative set and Theme-alternative set (Steedman, 2000) I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

43

IV

A VIE N

IS and Discourse Dynamics • IS and the File-Change Metaphor (Vallduv´ı, 1992) Theme : “ushers” hearer to a specific file-card address Rheme : provides information to add/modify on the card • IS-Sensitive Context Steedman, 2000)

Update

(Krifka, 1993;

θ (ψ ) c1


c2

ρ (ψ )

IS & Presentation

Kruijff-Korbayová, 1998;

c3


R

S

IS

S

UN

E R SIT A

SA

44

IV

A VIE N

Information Packaging (Chafe, 1976), (Vallduv´ı, 1992; Vallduv´ı, 1994), (Vallduv´ı and Engdahl, 1996) • IS-partitioning into Ground and Focus; Ground further partitioned into Link and Tail • partitioning defined on surface form, not on sentence meaning! • semantics of IP in terms of operations on file-cards: create, go-to, update (“file-change” metaphor taken literally) cf. also (Reinhart, 1995; Erteschik-Shir, 1997) • (Vallduv´ı and Engdahl, 1996): analysis of IP realization in many languages


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

45

IV

A VIE N

Vallduv´ı: IP Examples Link-Focus: (44) (45) (46) (47)

The boss [F called ]. The boss [F visited a broccoli plantation in colombia ]. The boss [F I wouldn’t bother ]. Broccoli the boss [F doesn’t eat ].

Link-Focus-Tail: (48) The boss [F hates ] broccoli. (49) The farmers [F already sent ] the broccoli to the boss.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

46

IV

A VIE N

Vallduv´ı: IP Examples All Focus: (50) [F The boss called ]. (51) Waiter! [F There’s a fly in my cream of broccoli soup ]! (52) What doesn’t the boss like? [F Broccoli ]. Focus-Tail: (53) I can’t believe this! The boss is going crazy! [F Broccoli ], he wants now.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

47

IV

A VIE N

IP and File Change Metaphor (Vallduv´ı, 1992) • operations on cards: – go to (introduce) a new card – go to an existing card – access a record on a card – add/modify a record on a card • four possible instruction types for IS: – update-add(I S ) for linkless all-focus sentence – update-replace(I S ,record(f c)) for focus-tail sentence – goto(f c),update-add(I S ) for link-focus sentence – goto(f c),update-replace(I S ,record((f c)) for link-focus-tail sentence I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

48

IV

A VIE N

Vallduv´ı: File Change Example(s) (54)

a. H: I’m arranging things for the president’s dinner. Anything I should know? b. S: Yes. The president [F hates the Delft china set ]. Don’t use it. c. goto(125) (update-add(hates the Delft-china-set(125))

(55)

a. H: In the Netherlands I got the president a big Delft china tray that matches the set he has in the living room. Was that a good idea? b. S: Nope. The president [F hates ] the Delft china set. c. goto(125) (update-replace(hates, { : Delft-china-set(125) }))


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

49

IV

A VIE N

Vallduv´ı: File Change Example(s) (56)

H: I’m arranging things for the president’s dinner. Anything I should know? S: Yes. The president always uses plastics dishes. [F (He) hates the Delft china set ]. update-add(hates the Delft-china-set(125))

(57)

H: In the Netherlands I got the president a big Delft china tray that matches the set he has in the living room. Will the president like it? S: Nope. [F (He) hates ] the Delft china set. update-replace(hates, { : Delft-china-set(125)})


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

50

IV

A VIE N

IS Semantics Using Alternative Sets • Semantics of IS in terms of selecting one member from a presupposed set of alternatives (Steedman, 2000) – Theme presupposes a Rheme-alternative set, i.e., a set of alternative propositions that could possibly answer the corresponding question in the given context; Rheme then restricts the Rheme-alternative set to a singleton – Theme also presupposes a Theme-alternative set, i.e. a set of alternative questions; Focus within Theme then restricts the Theme-alternative set to a singleton • These are pragmatic presuppositions that the relevant alternative set(s) be available in the context. • The systematic recognition of the alternative sets, and their maintenance as a discourse progresses are open research issues. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

51

IV

A VIE N

AS Semantics: Examples (58) I know how to transport babies in the metro. But what about dogs? must be carried Dogs L+H*LH% H*LL% {z } | {z } | {z } | F ocus | {z } | Background {z F ocus } T heme

Rheme

θ(58): λQ. Q (?dog 0) ρ(58): λx. ? carry 0(hearer 0, x) ρ-AS(58): {carry 0(hearer 0, dog 0 ), walk on lead(hearer 0, dog 0), load on buggy(hearer 0, dog 0 )} θ-AS(58): {∃Q.Q(dog 0), ∃P.P (baby 0)} I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

52

SA

cont’d

IV

A VIE N

AS Semantics: Examples (59) I know what must be worn in the metro. But what must be carried? Dogs must be carried H*LL% L+H*LH% | {z } | {z } | {z } {z } | Background {z F ocus | F ocus } Rheme

T heme

θ(59): λx. ? carry 0(hearer 0, x) ρ(59): λQ. Q (?dog 0) ρ-AS(59): {carry 0(hearer 0, dog 0 ), carry 0(hearer 0, baby 0)} θ-AS(59): {∃x.carry 0(hearer 0, x), ∃x.wear(hearer 0, x)}


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

53

IV

A VIE N

AS and Discourse Dynamics θ (ψ ) c1

c2

ρ (ψ )

c3

Theme update phase : c1 [θ(ψ)]c2 • verify Theme presuppositions ASθ (ψ) and ASρ(ψ); • restrict ASθ (ψ) Rheme update phase : c2 [ρ(ψ)]c3 • verify and restrict ASρ(ψ) Note: the “intermediate” context resulting from Theme-update available as a context for the interpretation of subsequent utterances, cf. IS-sensitive interpretation of “otherwise” (Kruijff-Korbayová and Webber, 2001). I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

54

IV

A VIE N

Updating with “otherwise β” (60) (What should I do at a red light?) α: At a red light, stop. H*LL% | {z }| {z } T heme

c0

β: Otherwise you will get a ticket.

Rheme

θ (α )

................. . . . . . . .... .... c1

ρ (α )

c2

ρ (α ) c3

θ (β )

c4

ρ (β )

c5

If the light is red and you do not stop, you will get a ticket.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

55

IV

A VIE N

Meaning Differences


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

56

IV

A VIE N

IS: Meaning Differences (61) Dogs must be carried. a. What about dogs?

(Halliday, 1970) Dogs must be carried. | {z } | {z } T heme

Rheme

i.e., If there is a dog, it must be carried. Presupposed alternative set: {∃P.(P (e) ∧ patient(e, dog))} b.

Dogs must be carried. | {z } Rheme

Presupposed alternative set: {∃P.P (e)}


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

57

IV

A VIE N

IS: Meaning Differences (62) Smoke in the hallway. a. Where should one smoke?

(Hajiˇcová, 1993) Smoke in the halllway. | {z } | {z } T heme

Rheme

i.e., If you (want to) smoke, do it in the hallway. Presupposed alternative set: {∃x.(smoke(e) ∧ location(e, x))} b. What should one do in the hallway? |

Smoke in the hallway. {z } | {z } Rheme

T heme

i.e., If you are in the hallway, smoke. Presupposed alternative set: {∃P.(P (e) ∧ location(e, hallway))}


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

58

IV

A VIE N

IS: Meaning Differences (63) Staff behind counter. a. Where should staff be?

(Hajiˇcová, 1993) Staff behind counter. | {z } | {z } T heme

Rheme

i.e. Where staff should be is (only) behind the counter. Presupposed alternative set: {∃x.location(staf f, x)} b. Who should be behind the counter? |

Staff behind counter. {z } | {z } Rheme

T heme

i.e., Who should be behind the counter is (only) staff. Presupposed alternative set: {∃y.location(y, behind counter)}


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

59

IV

A VIE N

IS: Meaning Differences (64)

a. What language is (mostly) spoken on the Shetlands? On the Shetlands one speaks English. | {z }| {z } T heme

Rheme

Presupposed alternative sets: ρ-AS: {∃x.(speak(e) ∧ location(e, shetlands) ∧ language(e, x))} θ-AS: {∃z.∃x.(speak(e) ∧ location(e, z) ∧ language(e, x))} b. Where is English (mostly) spoken? One speaks English on the Shetlands. {z }| {z } | T heme

Rheme

Presupposed alternative sets: ρ-AS: {∃y.(speak(e) ∧ language(e, english) ∧ location(e, y))} θ-AS: {∃z.∃y.(speak(e) ∧ language(e, z) ∧ location(e, y))} I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

60

IV

A VIE N

IS: Meaning Differences (65) Officers always escorted ballerinas. (Partee et al., 1998) a. Whom did officers always escort? Officers escorted always ballerinas. | {z } | {z } T heme

Rheme

{∃x.(escort(e) ∧ actor(e, of f icer) ∧ patient(e, x))} b. What did officers always do? Officers always escorted ballerinas. | {z } | {z } T heme

Rheme

{∃P.(P (e) ∧ actor(e, of f icer))} c. Who always escorted ballerinas? Officers always escorted ballerinas. {z } | {z } | Rheme

T heme

{∃y.(escort(e) ∧ actor(e, y) ∧ patient(e, ballerina))} I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

61

IV

A VIE N

IS: Meaning Differences Czech newspaper 1990:

(Hajiˇcová)

ˇ si udˇelali revoluci. (66) Dobrá zpráva je, ˇze Ceˇ ˇ ˇ si. Spatn´ a zpráva je, ˇze revoluci udˇelali Ceˇ The good news is that the Czechs made a revolution; the bad news is that a revolution was made by the Czechs. (. . . the bad news is that the Czechs made a revolution.) {∃x.(make(e) ∧ actor(e, czechs) ∧ patient(e, x))} {∃y.(make(e) ∧ actor(e, y) ∧ patient(e, revolution))}


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

62

IV

A VIE N

(67) Problém nen´ı v tom, ˇze Janouch koupil gamma n˚ uˇz, ale ˇze gamma n˚ uˇz koupil Janouch. The problem is not that Janouch bought a gamma-knife, but that the gamma-knife was bought by Janouch. (. . . but that Janouch bought the gamma knife.) {∃x.(buy(e) ∧ actor(e, janouch) ∧ patient(e, x))} {∃y.(buy(e) ∧ actor(e, y) ∧ patient(e, gknif e))}


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

63

IV

A VIE N

Intermezzo


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

64

IV

A VIE N

“Focussing jokes” (68)

Why do you rob banks? Because that’s where the money is!

(69)

Why do firemen wear red suspenders? To keep their pants up.

(70)

Why do we buy clothes? Because we can’t get them for free.

(71)

Why do we dress girls in pink and boys in blue? Because they can’t dress themselves.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

65

IV

A VIE N

IS in Computational Applications


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

66

IV

A VIE N

Computational Modelling of IS in Applications • word order: – anaphora resolution: Cz. (Hajiˇcová et al., 1990; Hajiˇcová et al., 1992) – database-querying dialogue: Turkish (Hoffman, 1995b) – MT: En.→Turkish (Hoffman, 1996); En.→Polish (Sty´s and Zemke, 1995) – text generation: Cz., En. (Kruijff-Korbayová et al., 2002) • intonation in question-answer pairs and in dialogue: En. (Prevost, 1995), Ger., En. (Kruijff-Korbayová et al., 2003); En. (Moore et al., 2004; Baker et al., 2004) • referring expression generation in text (Hajiˇcová et al., 1990) and in dialogue (Krahmer and Theune, 2002) • non-linguistic aspects in modeling embodied conversational agents: gestures (Pelachaud et al., 1998; Cassell et al., 2000), gaze and turn-taking (Cassell et al., 1999)


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

67

IV

A VIE N

IS in Answers to Questions


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

68

IV

A VIE N

IS-Based Word Order Modeling • Modeling contextually appropriate word order in Turkish in answers to wh- and yes/no-questions in a DB query task (Hoffman, 1995b) • Approach to IS Vallduv´ı, 1994):

based

on

information

packaging

(Vallduv´ı, 1992;

• Association of sentence positions with discourse functions: – sentence initial position tends to be the topic – immeditely preverbal position tends to be focus – elements between topic and focus and postverbal elements are in the ground • CCG-based grammar formalization I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

69

IV

A VIE N

IP Representation (Hoffman, 1995b; Hoffman, 1995a): topic vs. comment (=ground/focus) 2

(72)

3

syn: . . . 6 sem: . . . 7 6 2 3 7 6 7 topic: . . . » 6 7 – 6 5 7 focus: . . . 4 info: 4 5 comment: ground: . . .

• Topic has the value “recoverable” when zero-pronoun or in verb-initial sentences (all-focus) • T/C structures fully recursive


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

70

IV

A VIE N

IP Representation (Hoffman, 1995b): (73) D¨ un Fatma’nın gitti˘gini Ay¸se biliyor. Yesterday Fatma-Gen go-Ger-Acc Ay¸se knows. It’s Ays¸e who knows that yesterday, Fatma left. 3 syn: . . . 7 6 sem: . . . 7 6 3 2 3 2 7 6 topic: yesterday 7 6 » – 6 7 6 topic: 4 5 7 7 focus: Fatma 7 6 6 comment: 6 7 7 6 6 info: 6 ground: go 7 7 – » 6 7 7 6 4 focus: Ay¸ s e 5 5 4 comment: ground: know 2


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

71

IV

A VIE N

DB Question Answering System 1. Parser determines syn, sem, inf o 2. Planner executes simple plans to handle different types of questions: i. determine question type (sem : type): (a) wh-q; (b) yes/no-q: Prop-q (q-morph on verb); Focused-q (q-morph on non-verb); Schedule-q (ability) ii. query DB with sem : lf , respecting IP of question if success then generate corresponding answer else generate a “negative” answer iii. plan answer: copy as much as possible from question, add/modify IP: topic of question → topic of answer; info from DB → focus


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

72

IV

A VIE N

Hoffman: Example 1 (74) Fatma’yı kim aradı? Fatma-Acc who call-Past? As for Fatma, who called her? 2

3

syn: . . . 2 3 6 7 event: 7349 6 7 6 sem: 4 type: quest(lambda( 7350)) 7 5 6 7 6 7 lf: { call( 7349, 7350,fatma), . . . } 6 7 6 7 6 7 2 3 7 6 6 7 topic: person(fatma) » – 6 7 5 4 info: 4 5 focus: person( 7350) comment: ground: call( 7349, 7350,fatma) db file(fatma, person(fatma)). db file(fatma, call(e3,ayse,fatma)). db file(fatma, see(e4,fatma,ahmet)).


IS & Presentation


R

S

IS

S

UN

E R SIT A

73

SA

cont’d

IV

A VIE N

Hoffman Example 1 db file(fatma, person(fatma)). db file(fatma, call(e3,ayse,fatma)). db file(fatma, see(e4,fatma,ahmet)).

(75) Fatma’yı Ay¸se aradı. Fatma-Acc Ay¸se call-Past As for Fatma, it was Ay¸se who called her. 2

syn: . .». – event: e3 sem: lf: { call(e3,ayse,fatma), . . . }

3

7 6 7 6 7 6 7 6 7 6 6 3 7 2 7 6 7 6 topic: person(fatma) » – 6 5 7 5 4 info: 4 focus: person(ayse) comment: ground: call(e3,ayse,fatma)


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

74

IV

A VIE N

Hoffman: Example 2 db file(fatma, person(fatma)). db file(fatma, call(e3,ayse,fatma)). db file(fatma, see(e4,fatma,ahmet)).

(76) Fatma’yı Ahmet mi aradı? Fatma-Acc Ahmet Quest call-Past As for fatma, was it Ahmet who called her? 2

3

syn: . . . 2 3 6 7 event: 9041 6 7 6 sem: 4 type: quest(yes/no,ahmet) 7 5 6 7 6 7 lf: { call( 9041,ahmet,fatma), . . . } 6 2 3 7 6 7 6 7 topic: person(fatma) » – 6 5 7 4 info: 4 5 focus: person(ahmet) comment: ground: call( 9041,ahmet,fatma)


IS & Presentation


R

S

IS

S

UN

E R SIT A

75

SA

cont’d

IV

A VIE N

Hoffman: Example 2 db file(fatma, person(fatma)). db file(fatma, call(e3,ayse,fatma)). db file(fatma, see(e4,fatma,ahmet)).

(77) Hayır, Fatma’yı Ay¸se aradı. No, Fatma-Acc Ay¸se call-Past No, as for 2Fatma it was Ay¸se who called her.

syn: . .». – event: e3 sem: lf: { call(e3,ayse,fatma), . . . }

3

7 6 7 6 7 6 7 6 7 6 6 3 7 2 7 6 7 6 topic: person(fatma) » – 6 5 7 5 4 info: 4 focus: person(ayse) comment: ground: call(e3,ayse,fatma)


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

76

IV

A VIE N

IS-Based Assignment of Intonation • Using IS to control intonation of synthesized spoken output in answers to questions: question IS fully determines the answer IS (Prevost and Steedman, 1993) • Theme/Rheme determination: – rheme of the question determines the theme of the answer • Focus determination: – terms focused in question are focused in asnwer – term instantiating question variable is also focused – for more complex rhemes, only new elements are focused I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

77

IV

A VIE N

Prevost: Example (78) I know that widgets contain cogs, but what parts do wodgets include? L+H* LH% H* LL% prop: s : λx[part(x)&include(?wodgets, x)] theme: s : λx[part(x)&include(?wodgets, x)] / (s : include(?wodgets, x)/np : x) rheme: s : include(?wodgets, x)/np : x prop: s : include(?wodgets, ?sprockets) theme: s : include(?wodgets, x)/np : x) (79) rheme: np : ?sprockets Wodgets include sprockets. L+H* LH% H* LL% I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

78

IV

A VIE N

IS-Based Assignment of Intonation • (Prevost and Steedman, 1994): IS determination in answer not from question alone but also from discourse model (database) • Theme/Rheme determination: – rheme of the question determines the theme of the answer • Focus determination: – terms focused in question’s rheme are focused in answer’s theme – rheme-focus in answer determined from alternative sets in the database


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

79

IV

A VIE N

Constructing Rheme-Alternative Sets (Prevost and Steedman, 1994): Given database D, object x and a set of properties P that uniquely describe x: 1. construct a set of objects, A, (and their referring properties) which can be considered alternatives to x w.r.t. D 2. restrict A by properties of objects mentioned in theme → A0 3. mark as contrastive those properties of x in P that exclude some alternatives from A0


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

80

IV

A VIE N

IS in Answers to Questions: Summary • IS and WO (Hoffman); IS and intonation (Prevost); formalization in CCG • Theme/Rheme determined by IS in question (differences in IS assignment in questions between Hoffman and Prevost!) • Focus is determined from (a) question, (b) contrast w.r.t. alternatives in discourse model (database) • Hoffman’s database contains propositions and is organized by themes (entities) • Prevost’s discourse model (database) contains entities, propositions, and themes and rhemes I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

81

IV

A VIE N

IS in Dialogue Systems


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

82

IV

A VIE N

Intonation of Spoken Dialog System Output (Kruijff-Korbayová et al., 2003) • Most current systems have limited dialogue flexibility, which enables them to use carefully scripted interactions with predefined and prerecorded output • However, flexible interaction requires output to be dynamically generated • The realization of dynamically generated output needs to be controlled, to ensure that it is contextually appropriate • In particular, the intonation of synthesized spoken output needs to be controlled w.r.t. the context


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

83

IV

A VIE N

Motivation: Intonation in context U: Which devices are in the house? S:

There is a stove H* and a radio

in the kitchen, H* LH% in the bathroom. H* LL%

a radio H*

in the kitchen LH%

U: What is the status of the devices in the kitchen? S:

The stove L+H* The radio L+H*


in the kitchen L% in the kitchen L%

is on. H*LL% is off. H*LL%

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

84

IV

A VIE N

Information-State Update Approach to Dialogue Modeling • Dialog moves are modeled as information state update transitions • Information State represents the current discourse context (in a dialogue participant’s view) • e.g. a version of the Dialogue Game Board (Ginzburg, 1996) in GoDIS:


IS & Presentation


:

4 2 6 6 4

lu

:

: : :

Stack(Action) StackSet(Action) 5 Set(Proposition) 3 Set(Proposition) 7 Stack(Question) » – 7 5 speaker : Participant moves : AssocSet(Move,Bool)

IS & Presentation

R

S

IS

UN


:

agenda plan bel com : qud :

3

S

6 private 6 6 6 6 6 6 6 shared 4

2

E R SIT A

2

SA

85

IV

A VIE N

3 7 7 7 7 7 7 7 7 5


R

S

IS

S

UN

E R SIT A

SA

86

IV

A VIE N

Information Structure Assignment Theme/Rheme partitioning: determined according to the QUD • QudTR rule: given an utterance content u to partition, if QUD corresponds to the result of λ-abstracting over a part of u, this part is marked as the Rheme If on QUD: ?λx.status(x), then status( on ) | {z } Rheme


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

87

IV

A VIE N

Information Structure Assignment Background/Focus partitioning: determined by comparing parallel elements • ComFB rule: if there is an element in the shared commitments that is parallel but not identical to an element in the utterance content, the part that is non-identical is marked as the Focus If in shared commitments: {type(stove)&location(kitchen); type(radio)&location(kitchen); type(radio)&location(bathroom)} then type( stove ) & location(kitchen) | {z } F ocus


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

88

IV

A VIE N

Information Structure Assignment Background/Focus partitioning: determined by comparing parallel elements • ComFB rule: if there is an element in the shared commitments that is parallel but not identical to an element in the utterance content, the part that is non-identical is marked as the Focus • DomFB rule: if there is an element in the domain model that is parallel but not identical to an element in the utterance content, the part that is non-identical is marked as the Focus


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

89

IV

A VIE N

Information Structure Assignment Propositional content

QudTR

TR−partitioned propositional content

ComFB

DomFB

IS partitioned propositional content


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

90

IV

A VIE N

Experimental Implementation Producing synthesized output with contextually varied intonation in GoDiS IS−Partitioning

Generation module

Output module

Text interface

Festival interface

MARY interface

Text

SABLE/ AMPL

SABLE/ MaryXML

Festival

MARY

Text output

Audio output


Audio output

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

91

IV

A VIE N

Experimental Implementation Evaluation • Using the German text to speech synthesis system Mary (Schr¨ oder and Trouvain, 2001) which supports intonation annotation using GToBI (Grice et al., to appear) • Experiment 1: default vs. controlled intonation using GToBI or SABLE – Dialogue fragments displayed on screen – Several turns provide context for target utterance – Target utterance synthesized in different versions – Subjects judge appropriateness of intonation in the given context • Experiment 2: only default vs. GToBI controlled intonation – Subjects judge intonation without context – Subjects judge appropriateness of intonation in the given context I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

92

IV

A VIE N

Experimental Evaluation Results Although the results are not significant, observed tendencies correspond to expectations: • overall average judgments worse for default than for controlled intonation • average judgments per IS pattern also worse for default than for controlled intonation (not much difference across patterns, though one would expect it!) • judgments of default intonation in isolation closer to those where the context is matching with this, then to those where the context does not match • roughly same results whether looking at absolute values of judgments or taking differences between values in isolation and in context, per subject I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

93

IV

A VIE N

Experimental Evaluation Experience • Proper (standard) evaluation methodology is lacking • Indirect evaluation through task success / completion time does not seem suitable, because of accumulation of effects through dialogue (moreover, it would have to be Wizard of Oz, because of coverage and robustness issues) • Direct evaluation is hard to design as a proper experiment: – Do subjects really take context into account? – Are they judging contextual appropriateness of the intonation pattern and not the quality of the synthesized output as such? ∗ Absolute judgments allow comparison of judgments across dialogues ∗ Comparative judgments could neutralize synthesis quality I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

94

IV

A VIE N

Summary • Domain- and application-independent rules determining IS partitioning from the information state • Domain- and application-independent rules mapping IS partitioning to realization through intonation (in template-based generation) • Experimental implementation using TTS systems which support ToBI-based intonation determination • Test-of-concept evaluation


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

95

IV

A VIE N

FLIGHTS • Fancy Linguistically Informed Generation of Highly Tailored Speech (Moore et al., 2004; Baker et al., 2004) – tailored descriptions of the most relevant available options and their attributes in information-seeking dialogue (flight information) – content selection and contextually appropriate information presentation in order to make descriptions easy to understand and memorable: ∗ taking user preferences into account when deciding about various aspects: referring expressions, aggregation, discourse cues, scalar terms ∗ contextually appropriate intonation • Approach to IS based on (Steedman, 2000)


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

96

IV

A VIE N

FLIGHTS: IS • content planner determines theme/rheme and focus/background • OpenCCG realizer makes intonation choices • Intonation markup for TTS using APML, an XML markup for specifying turn-taking, performative, affective and informational aspects of text • TTS with Festival cluster unit synthesis, using restrictions on unit types which reflect pitch accents and boundary tones, cf. ToBI (Silverman et al., 1992) • Evaluation by perceptual comparison between different versions, on descriptive sentences and clarification requests (significant results) I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

97

IV

A VIE N

IS in MT and Text Generation


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

98

IV

A VIE N

Target WO in English → Turkish MT (Hoffman, 1996) • Determination of Topic and Focus in target w.r.t. contextual information. • Not using cues from source language text! • Using centering, old/new and contrastiveness. • Topic and Focus determined by algorithms; the rest is Ground.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

99

IV

A VIE N

Topic Determination Algorithm Given: • sentence contents, • list of discourse entities mentioned in text so far, • C f lists of current and preceding sentence (cf. Centering (Grosz et al., 1995)) Topic determination: 1. Try to choose most salient discourse-old entity. 2. Else try to choose a situation-setting adverb. 3. Else choose the first item on the C f list of current sentence (i.e., Subject) I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

100

IV

A VIE N

Focus Determination Algorithm Given: • the non-topic rest of the sentence contents, • list of discourse entities mentioned in text so far, Focus determination: 1. If there are any discourse-new entities, put them into focus. 2. Else determine contrastive focusing of discourse-old information: For each entity: i. Construct a set of alternatives based on the entity’s semantic type ii. If the alternative set is not empty, put the entity into focus I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

101

IV

A VIE N

Target WO in Polish → English MT Contrary to (Hoffman, 1996), (Sty´s and Zemke, 1995) argue for discourse analysis of the source text in order to preserve its communicative meaning in MT. • Tracking centers according to Centering Theory (Grosz et al., 1995) • Additional criteria for center evaluation: special center-pointing constructions, demonstrative pronouns, possessive and demonstrative modifiers, definiteness award, indefiniteness penalty • Further modifications: gradation of center values, center values for all NPs, composite computation of center values, referential distance, synonyms • Set of ordering criteria (end weight, given fronting, short before long, specific patterns) and preferences based on statistical models I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

102

SA

Background

IV

A VIE N

Centering Theory (Grosz et al., 1995) • Attempt to – account for attentional limitations of discourse participants: they can only attend to a small number of referents at the same time – reduce inference load in the process of discourse interpretation (i.e., more likely candidates for coreference considered first) • There are local and global aspects of attention centering (e.g., based on overall task structure or communicative goals) • CT is a computational model of local centering of attention I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

103

SA

Background

IV

A VIE N

Centering Theory (80)

(81)

a. b. c. d. a. b. c. d.

John went to his favorite music store to buy a piano. He had frequented the store for many years. He was excited that he could finally buy a piano. He arrived just as the store was closing. John went to his favorite music store to buy a piano. It was a store John had frequented for many years. He was excited that he could finally buy a piano. It was closing just as John arrived.


IS & Presentation


R

S

IS

S

UN

E R SIT A

104

SA

Background

IV

A VIE N

Centering Theory • each utterance has one backward looking center C b and an ordered set of forward looking centers C f • proposed C f ordering Subj < Obj < Other (various other proposals considered in the literature) • the most highly ranked item on C f is the C p , i.e., the preferred C b for the next utterance • Rule 1: If any item is pronominalized, then the C b is pronominalized • types of center-transitions depending on whether backward looking center is maintained or changed: continuation, retaining, shift • Rule 2: preference for sequences of center continuation, or smooth (=gradual) shift


IS & Presentation


R

S

IS

S

UN

E R SIT A

105

SA

Background

IV

A VIE N

Centering Theory Centering transition types:

C b (U i ) = C p (U i ) C b (U i ) 6= C p (U i )

C b (U i ) = C b (U i −1 ) Continue Retain

C b (U i ) 6= C b (U i −1 ) Smooth Shift Rough Shift

C b (U k ) — backward looking center of utterance U k C f (U k ) — (partially) ordered list of forward looking centers of utterance U k C p (U k ) — highest ranked item on C f (U k ), the preferred (next) center


IS & Presentation


R

S

IS

S

UN

E R SIT A

106

SA

Background

IV

A VIE N

Centering theory (82)

a. b. c. d. e.

The sentry was not dead. He was, in fact, showing signs of reviving. He was partially uniformed in a cavalry tunic. Mike stripped this from him and donned it. He tied and gagged the man. (Strube and Hahn, 1999)


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

107

IV

A VIE N

WO in Czech and English Intructional Texts • Flexible context-dependent WO in Cz and En in procedural instructions in software manual texts (Kruijff-Korbayová et al., 2002) • Several contextual factors determined by text and sentence planning: – Status of entities referred to: familiarity and identifiability – Information structure: topic-focus articulation (Sgall et al., 1986) – Discourse organization: genre-specific thematic scaffolding; theme as “point of departure” (Halliday, 1967; Halliday, 1985) • Language-specific constraints imposed by grammaticality and defaults (unmarked ordering) • Realization by a systemic functional grammar in KPML; general WO algorithm I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

108

IV

A VIE N

Examples of WO Factors Familiarity and indetifiability: (83) Otevˇrete nový soubor. Nakreslete kruh. Open a new file. Draw a circle. a. Soubor uloˇzte. File save. Save the file. b. # Uloˇzte soubor. Save file.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

109

IV

A VIE N

Examples of WO Factors Information structure: (84) Obrázek obsahuje kruh a obdéln´ık. The drawing contains a circle and a rectangle. a. Kruh vymaˇzte. Circle delete. b. Vymaˇzte kruh. Delete circle. Delete the circle.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

110

IV

A VIE N

Examples of WO Factors Theme as point of departure: (85)

a. In the File menu choose the Open command. b. Choose the Open command in the File menu.

(86)

a. By pressing Ctrl-O open a file. b. Open a file by pressing Ctrl-O.

(87)

a. To open a file press Ctrl-O. b. Press Ctrl-O to open a file.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

111

IV

A VIE N

IS-Based Assignment of Intonation in Text • Previous work: accenting affected by “givenness” (Hirschberg 1990), (Hirschberg, 1993) • (Prevost, 1996): – accenting based on (a) old/new; (b) contrast – text organization takes thematic progression into account (prefers theme continuation)


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

112

IV

A VIE N

IS-Based Assignment of Intonation • Content and text planning: determine a sequence of propositions about an object and the rhetorical relations, segment each proposition into theme/rheme – discourse model contains previous themes and rhemes (ISstore) – to determine theme, search for most recent match, prefer theme-continuation – determine rheme as complement of theme • Sentence planning: determination of realization, focus assignment – each new (not mentioned) property or discourse entity get focus – contrasting elements get focus


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

113

IV

A VIE N

Assigning Contrastive Focus Step 1: Determine contrasting propositions • containing 2 contrasting pairs of entities or 1 pair of contrasting entities and contrasting functors • discourse entities are contrasting when they are alternatives w.r.t. isa in DB Step 2: Contrastive focus algorithm Given: object x, properties P , alternatives A: 1. restrict A to objects mentioned in discourse → A0 2. for each property p in P , include p in set of contrasting properties P c iff p excludes some object from A0 I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

114

IV

A VIE N

IS-Based Intonation in TTS • Previous work: acenting affected by “givenness” (Hirschberg 1990), (Hirschberg, 1993) • (Hiyakumoto et al. 1997): – combine first mention and contrastiveness as reasons for accenting – use of WordNet in givenness and contrast determination: to identify sets of synonyms and contrasting words for open-class words (nouns, verbs, adjectives, adverbs) – determine theme/rheme in propositional constituents by heuristics applied to pre- and post-verbal and verb-complex material, and considering presence of focus within it


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

115

IV

A VIE N

Summary: IS in MT, Text Generation and TTS • IS and WO; IS and intonation • Theme/Rheme determined by (a) linking in text (centering; continuation); (b) genre-specific thematic scaffolding; (c) heuristics

theme

• Focus is determined by (a) discourse newness, (b) contrast w.r.t. alternatives in discourse model • Combination of contextual and grammatical factors (CCG, SFG)


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

116

IV

A VIE N

IS in Multimodal Interaction


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

117

IV

A VIE N

IS in Multimodal Interaction • Appropriate and synchronized speech, intonation, facial expressions and hand gestures (Pelachaud et al., 1998) • Integrating turn-taking and IS provides better explanation for gaze behavior (Cassell et al., 1999) • Generation of either speech, gesture or combination of both as a function of IS status and surprise value of a discourse entity (Cassell et al., 2000) • Various researches have observed that distance, posture shifts and other body movements seem to accompany changes in the topic or social relationship


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

118

IV

A VIE N

IS and Gestures Gilbert and George (Pelachaud et al., 1998); Greta ECA (http://www.iut.univ-paris8.fr/greta/) • Some facial expressions are automatically generated according to intonation (e.g., raising eyebrows with accented material), based on corpus studies • Head nods and look-toward listener punctuate accented and emphasized items • Iconic and metaphoric gestures (i.e., representing something) are generated for – rhematic verbal elements (roughly, information not yet spoken about) – hearer new references provided that the semantic content can receive such a gesture (e.g., spatial) • Beat gestures are generated – otherwise – to accompany discourse new definite references • Duration of intonation phrases is used to time gestures I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

119

IV

A VIE N

Wrap Up


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

120

IV

A VIE N

Summary and Conclusions • Two dimensions of IS-partitioning: Theme-Rheme, Focus-Background • IS realization means: intonation, word order, syntactic constructions, ... • IS is an essential aspect of meaning at the interface between utterance and discourse • IS is important for accurate NL processing • Formal accounts are emerging, some embodied into practical systems • many questions concerning IS partitioning and its realization in different languages still open • Further research topics: (Kruijff-Korbayová and Steedman, 2003) – further systematization of terminologies – formalization and computational modeling&testing – empirical and corpus-based studies – cross-linguistic investigations and multilingual applications I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

121

IV

A VIE N

That’s it folks! Questions?


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

122

IV

A VIE N

References Rachel Baker, Robert A. J. Clark, and Michael White. 2004. Synthetizing contextually appropriate intonation in limited domains. In Proceedings of SSWS. Dwight Bolinger. 1965. Forms of English. Harvard University Press, Cambridge, MA. Gillian Brown. 1983. Prosodic structure and the given/new distinction. In Anne Cutler, D. Robert Ladd, and Gillian Brown, editors, Prosody: Models and Measurements, pages 67–77. Springer-Verlag, Berlin. Daniel B¨ uring. 1995. The 59th Street Bridge Accent: On the Meaning of Topic and Focus. Ph.D. thesis, Universität T¨ ubingen. Publ. as (B¨ uring, 1997). Daniel B¨ uring. 1997. The Meaning of Topic and Focus: The 59th Street Bridge Accent. Routledge, London. Daniel B¨ uring. 1999. Topic. In Peter Bosch and Rob van der Sandt, editors, Focus: Linguistic, Cognitive and Computational Principles, Natural Language Processing, pages 142–165. Cambridge University Press, Cambridge. Justine Cassell, Obed. E. Torres, and Scott Prevost. 1999. Turn taking vs. discourse structure: How best to model multimodal conversation. In Y. Wilks, editor, Machine Conversations. Kluwer. Justine Cassell, Matthew Stone, and Hao Yan. 2000. Coordination and context-dependence in the generation of embodied conversation. In Proceedings of the INLG Conference. Wallace Chafe. 1976. Givenness, contrastiveness, definiteness, subjects, topics and points of view. In Charles Li, editor, Subject and Topic, pages 25–55. Academic Press, New Tork.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

123

IV

A VIE N

Noam Chomsky. 1965. Aspects of the Theory of Syntax. MIT Press, Cambridge, MA. Noam Chomsky. 1971. Deep structure, surface structure, and semantic interpretation. In Danny Steinberg and Leon Jakobovits, editors, Semantics, pages 183–216. Cambridge University Press, Cambridge. Herbert Clark and Susan Haviland. 1977. Comprehension and the given-new contract. In R. Freedle, editor, Discourse Production and Comprehension, pages 1–40. Ablex, Norwood, NJ. M.J. Cresswell. 1973. Logics and Languages. Methuen, London. ¨ Osten Dahl. 1969. Topic and Focus: a Study in Russian and General Transformational Grammar. Elandres Botryckeri, Göteborg. Nomi Erteschik-Shir. 1997. The dynamics of focus structure. Cambridge Studies in Linguistics. Cambridge University Press, Cambridge. Caroline Fery. 1993. German Intonational Patterns. Tuebingen:Niemeyer. Jan Firbas. 1964. On defining the theme in functional sentence analysis. Travaux Linguistiques de Prague, 1:267–280. Jan Firbas. 1966. Non-thematic subjects in contemporary english. Travaux Linguistiques de Prague, 2:229–236. Jan Firbas. 1971. On the concept of communicative dynamism in the theory of functional sentence perspective. Brno Studies of English, (7). Jan Firbas. 1992. Functional Sentence Perspective in Written and Spoken Communication. Studies in English Language. Cambridge University Press, Cambridge. Jonathan Ginzburg. 1996. Interrogatives: Questions, facts and dialogue. In Shalom Lappin, editor, The Handbook of Contemporary Semantic Theory, chapter Chapter 15, pages 385–422. Blackwell Publishers. M. Grice, S. Baumann, and R. Benzm¨ uller. to appear. German intonation in autosegmental phonology. In S.-A. Jun, editor, Prosodic typology. OUP. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

124

IV

A VIE N

H. P. Grice. 1975. Logic and conversation. In P. Cole and J. Morgan, editors, Syntax and semantics, number 3. Academic Press, New York. Joseph Grimes. 1975. The Thread of Discourse. Mouton. Barbara Grosz and Candace Sidner. 1986. Attention, intention and the structure of discourse. Computational Linguistics, 12:175–204. Barbara Grosz, Aravind Joshi, and Scott Weinstein. 1995. Centering: a framework for modeling the local coherence of discourse. Computational linguistics, 21(2):203–226. Janet Gundel. 1974. The Role of Topic and Comment in Linguistic Theory. Ph.D. thesis, University of Texas, Austin. Eva Hajiˇcová, Petr Kuboˇ n, and Vladislav Kuboˇ n. 1990. Hierarchy of salience and discourse analysis and production. pages 144–148. Eva Hajiˇcová, Vladislav Kuboˇ n, and Petr Kuboˇ n. 1992. Stock of shared knowledge - a tool for solving pronominal anaphora. pages 127–133. Eva Hajiˇcová. 1993. Issues of sentence structure and discourse patterns, volume 2 of Theoretical and computational linguistics. Charles University, Prague, Czech Republic. Michael Halliday and Ruqaiya Hasan. 1976. Cohesion in English. Longman. Michael A. K. Halliday. 1967. Notes on transitivity and theme in English — parts 1 and 2. Journal of Linguistics, 3(1 and 2):37–81 and 199–244. Michael A.K. Halliday. 1970. A Course in Spoken English: Intonation. Oxford Uniersity Press, Oxford. Michael A.K. Halliday. 1985. Introduction to Functional Grammar. Edward Arnold, London, U.K. C. Hamblin. 1973. Questions in Montague english. Foundations of Language, pages 41–53.


IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

125

IV

A VIE N

Irene Heim. 1982. The semantics of definites and indefinite noun phrases in English. Ph.D. thesis, University of Massachussetts, Amherst. H. Hendriks. 1998. Links without locations. In F. Hamm and E. Zimmermann, editors, Linguistische Berichte, Sonderheft 9/1998: SEMANTIK. Westdeutscher Verlag, Opladen. Julia Hirschberg. 1993. Pitch accent in context: Predicting intonational prominence from text. Artificial Intelligence, (63):305–340. Beryl Hoffman. 1995a. Computational Analysis of the Syntax and Interpretation of ”Free” Word-Order in Turkish. Ph.D. thesis, University of Pennsylvania. IRCS Report 95-17. Beryl Hoffman. 1995b. Integrating free word order, syntax, and information structure. In Proceedings of the 7th Conference of the European Chapter of the Association for Computational Linguistics, Dublin, pages 245–252, San Francisco, CA. Morgan Kaufmann. Beryl Hoffman. 1996. Translating into free word order languages. In Proceedings of the International Conference on Computational Linguistics (COLING-96), Copenhagen, pages 556–561, Copenhagen, Denmark. Center for Sprogteknologi. Ray Jackendoff. 1972. Semantic Interpretation in Generative Grammar. MIT Press, Cambridge, MA. Ray Jackendoff. 1990. Semantic Structures. MIT Press, Cambridge. Hans Kamp. 1984. A theory of truth and semantic representation. In Jeroen Groenendijk, Theo Janssen, and Martin Stokhof, editors, Truth, Interpretation, and Information, pages 1–41. Foris, Dordrecht. Lauri Karttunen and Stanley Peters. 1979. Conventional implicature. In Choon-Kyu Oh and David Dinneen, editors, Syntax and Semantics 11: Presupposition, pages 1–56. Academic Press, New York. Lauri Karttunen. 1968. What makes definite noun phrases definite? Technical report, Rand Corporation. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

126

IV

A VIE N

Martin Kay. 1975. Syntactic processing and functional sentence perspective. In Roger Schank and Bonnie NashWebber, editors, Theoretical Issues in Natural Language Processing, (Supplement), pages 12–15, MIT, June. Nobo Komagata. 2003. ??? Journal of Logic, Language and Information: Special Issue on Discourse and Information Structure, 12(3):??? Emiel Krahmer and Mariet Theune. 2002. Efficient context-sensitive generation of referring expressions. In van Deemter and Kibble (van Deemter and Kibble, 2002), pages 223–264. Angelika Kratzer. 1991. The representation of focus. In Arnim von Stechow and Dieter Wunderlich, editors, Semantics: An International Handbook of Contemporary Research, pages 825–834. de Gruyter, Berlin. Manfred Krifka. 1991. A compositional semantics for multiple focus constructions. Cornell Working Papers in Linguistics, 10:127–158. Manfred Krifka. 1992. A compositional semantics for multiple focus constructions. In J. Jacobs, editor, Informationsstruktur und Grammatik, Linguistische Berichte, Sonderheft 4, pages 17–53. Manfred Krifka. 1993. Focus and presupposition in dynamic semantics. Theoretical Linguistics, 19:269–300. Ivana Kruijff-Korbayová and Mark Steedman, editors. 2001. Information Structure, Discourse Structure and Discourse Semantics, ESSLLI2001 Workshop Proceedings, Helsinki, Finland, August 20-24. European Summer School in Logic, Language and Information (ESSLLI), The University of Helsinki. Ivana Kruijff-Korbayová and Mark Steedman. 2003. Discourse and information structure. Journal of Logic, Language and Information: Special Issue on Discourse and Information Structure, 12(3):249–259. Ivana Kruijff-Korbayová and Bonnie Webber. 2001. Information structure and the semantics of “otherwise”. In Kruijff-Korbayová and Steedman (Kruijff-Korbayová and Steedman, 2001), pages 61–78. Ivana Kruijff-Korbayová, John Bateman, and Geert-Jan M. Kruijff. 2002. Generation of contextually appropriate I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

127

IV

A VIE N

word order. In van Deemter and Kibble (van Deemter and Kibble, 2002), pages 193–222. Ivana Kruijff-Korbayová, Stina Ericsson, Kepa Joseba Rodr´ıguez, and Elena Karagjosova. 2003. Producing contextually appropriate intonation is an information-state based dialogue system. In Proceedings of the 10th Conference of the European Chapter of the Association for Computational Linguistics (EACL), pages 227–234. ACL. Ivana Kruijff-Korbayová. 1998. The Dynamic Potential of Topic and Focus: A Praguian Discourse Representation Theory. unpublished Ph.D. thesis, Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic. Geert-Jan M. Kruijff. 2001. A Categorial-Modal Logical Architecture of Informativity: Dependency Grammar Logic ´ and Information Structure. Ph.d. dissertation, Institute of Formal and Applied Linguistics (UFAL), Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic. William Mann and Sandra Thompson. 1987. Rhetorical structure theory: A framework for the analysis of texts. Technical Report RS-87-185, Information Science Institute, Marina del Rey, CA. Vilem Mathesius. 1929. Functional linguistics. In Josef Vachek, editor, Praguiana: Some Basic and Less Well-known Aspects of the Prague Linguistics School, pages 121–142. John Benjamins, Amsterdam. publ. 1983. Richard Montague. 1973. The proper treatment of quantification in ordinary english. In Jaakko Hintikka, J. M. E. Moravcsik, and Patrick Suppes, editors, Approaches to Natural Language: Proceedings of the 1970 Stanford Workshop on Grammar and Semantics, pages 221–242. Reidel, Dordrecht. Reprinted in Montague 1974, 247-279. Johanna Moore, Mary Ellen Foster, Oliver Lemon, and Michael White. 2004. Generating tailored, comparative descriptions in spoken dialogue. In Proceedings of the Florida Artificial Intelligence Research Symposium (FLAIRS). Barbara H. Partee, Eva Hajiˇcová, and Petr Sgall. 1998. Topic-Focus Articulation, Tripartite Structures, and Semantic Content. Kluwer, Amsterdam. Catherine Pelachaud, Justine Cassel, Norman Badler, Mark Steedman, Scott Prevost, and Mathew Stone. 1998. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

128

IV

A VIE N

Synthesizing cooperative conversation. In Harry Bunt and R. J. Beun, editors, Multimodal Human-Computer Communication - Systems, Techniques and Experiments, Springer Notes in Artificial Intelligence 1374, pages 68–88. Springer Verlag. Janet Pierrehumbert and Julia Hirschberg. 1990. The meaning of intonational contours in the interpretation of discourse. In Philip Cohen, Jerry Morgan, and Martha Pollack, editors, Intentions in Communication, pages 271–312. MIT Press, Cambridge, MA. Livia Polanyi and Remko Scha. 1983. The syntax of discourse. Text, 3:261–270. Scott Prevost and Mark Steedman. 1993. Generating contextually appropriate intonation. In Proceedings of the Sixth Conference of the European Chapter of the Association for Computational Linguistics, Utrecht, pages 332–340, San Francisco, CA. Morgan Kaufmann. Scott Prevost and Mark Steedman. 1994. Specifying intonation from context for speech synthesis. Speech Communication, 15:139–153. Scott Prevost. 1995. A Semantics of Contrast and Information Structure for Specifying Intonation in Spoken Language Generation. Ph.D. dissertation, IRCS TR 96-01, University of Pennsylvania, Philadelphia. Scott Prevost. 1996. An information structural approach to spoken language generation. In Proceedings of 34th Annual Meeting of the Association for Computational Linguistics, Santa Cruz. Ellen Prince. 1978. A comparison of it-clefts and wh-clefts in discourse. Language, (54):883–906. Tania Reinhart. 1995. Topics and the conceptual interface. In Hans Kamp and Barbara Partee, editors, Proceedings of the workshop on Conetxt-Dependence in the analysis of linguistic meaning, in Prague, February 1995, and in Bad Teinach, May 1995, volume 1, pages 355–393, Azenbacher st. 12, Stuttgart, Germany. Institut f¨ ur Maschinelle Sprachverarbeitung, Universität Stuttgart. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

129

IV

A VIE N

Mats Rooth. 1985. Association with Focus. Ph.D. thesis, University of Massachusetts, Amherst. Mats Rooth. 1992. A theory of focus interpretation. Natural Language Semantics, 1:75–116. Bertrand Russell. 1905. On denoting. Mind, 14:479–493. Harvey Sacks, Emanuel Schegloff, and Gail Jefferson. 1974. A simplest semantics for the organization of turn–taking for conversation. Language, 50:696–735. Marc Schröder and J¨ urgen Trouvain. 2001. The german text-to-speech synthesis system MARY: A tool for research, development and teaching. In The Proceedings of the 4th ISCA Workshop on Speech Synthesis, Blair Atholl, Scotland. Elisabeth Selkirk. 1984. Phonology and Syntax. MIT Press, Cambridge, MA. Petr Sgall, Eva Hajiˇcová, and Jarmila Panevová. 1986. The meaning of the sentence in its semantic and pragmatic aspects. Reidel, Dordrecht, The Netherlands. Kim Silverman, Mary Beckman, John Pitrelli, Marie Ostendorf, Colin Wightman, Patti Price, Janet Pierrehumbert, and Julia Hirschberg. 1992. Tobi: A standard for labeling english prosody. In Proceedings of the International Conference on Spoken Language Processing, Banff, Alberta, pages 867–870, Edmonton. University of Alberta. Mark Steedman and Ivana Kruijff-Korbayová. 2001. Introduction: Two dimensions of information structure in relation to discourse structure and discourse semantics. In Kruijff-Korbayová and Steedman (Kruijff-Korbayová and Steedman, 2001), pages 1–6. Mark Steedman. 1991. Structure and intonation. Language, 67:262–296. Mark Steedman. 2000. Information structure and the syntax-phonology interface. Linguistic Inquiry, 31(4):649–689. Peter Strawson. 1950. On referring. Mind, 59:320–344. Peter Strawson. 1954. A reply to mr. sellars. Philosophical Review, 63:216–231. I.Kruijff-Korbayová

IS & Presentation


R

S

IS

S

UN

E R SIT A

SA

130

IV

A VIE N

Michael Strube and Udo Hahn. 1999. Functional centering: Grounding referential coherence in information structure. Computational Linguistics. url. Malgorzata E. Sty´s and Stefan S. Zemke. 1995. Incorporating discourse aspects in english-polish mt: Towards robust implementation. In Proceedings of the RANLP Conference, Velingrad, Bulgaria. Susanne Uhmann. 1991. Fokusphonologie. Tuebingen:Niemeyer. Enric Vallduv´ı and Elisabet Engdahl. 1996. The linguistic realisation of information packaging. Linguistics, 34:459– 519. Enric Vallduv´ı and Maria Vilkuna. 1998. On rheme and kontrast. In Peter Culicover and Louise McNally, editors, Syntax and Semantics, Vol. 29: The Limits of Syntax, pages 79–108. Academic Press, San Diego, CA. Enric Vallduv´ı. 1990. The Information Component. Ph.D. thesis, University of Pennsylvania. Enric Vallduv´ı. 1992. The Informational Component. Garland, New York. Enric Vallduv´ı. 1994. The dynamics of information packaging. In Elisabeth Engdahl, editor, Integrating Infromation Structure into Constraint-based and Categorial Approaches, pages 3–26. ILLC/Department of Philosophy, University of Amsterdam, The Netherlands. Kees van Deemter and Rodger Kibble, editors. 2002. volume 143 of Lecture Notes. CSLI. Arnim von Stechow. 1990. Categorial grammar and linguistic theory. Studies in Language, 14:433–478. Arnim von Stechow. 1991. Focusing and backgrounding operators. In Werner Abraham, editor, Discourse Particles: Descriptive and Theoretical Investigations on the Logical, Syntactic and Pragmatic Properties of Discourse Particles in German, pages 37–84. John Benjamins, Amsterdam. Bonnie Webber. 1988. Tense as discourse anaphor. Computational Linguistics, 14:61–73. Terry Winograd. 1972. Understanding Natural Language. Edinburgh University Press, Edinburgh. I.Kruijff-Korbayová

IS & Presentation


Information Structure and its Use for Context-Sensitive

Information Structure and its Use for Context-Sensitive

Suggest Documents

Capital and Its Structure

SUPPLEMENTARY INFORMATION FOR: Structure

Crustal velocity structure of western India and its use - IISc

Korean Nominalizer kes and Its Information Structure ... - CiteSeerX

Information Retrieval, Information Structure, and Information Agents

Information Content Inclusion Relation and its Use in Database Queries

Factors Influencing Strategic Use of Information Technology and Its ...

The scientific information provided through patents and its limited use ...

Information and its use by animals in evolutionary ecology

E-textbook use, information seeking behaviour and its ...

INFORMATION TECHNOLOGY AND ITS USE IN THE CLASSROOM ...

Information Structure, Labor Contracts and the Strategic Use of Debt!

Green Information Technology and its Implication for

Capital and Its Structure

Quasometry, Its Use and Purpose

LINKING INFORMATION FOR MOBILE USE

Discourse and Information Structure - CiteSeerX

Mutual Visibility and Information Structure

Prediction, Information Structure, and Good

Information Structure and Discourse Modelling

Objects and information structure

Quantum Information and Spacetime Structure

Use of Information Technology and Information ...

The Oral Health Information Suite (OHIS): Its Use ... - Semantic Scholar