A (Hybrid) Logical Approach to Frame...

A (Hybrid) Logical Approach to Frame Semantics

Sylvain Pogodalla1

(joint work with Laura Kallmeyer2 and Rainer Osswald2)

1LORIA–INRIA, Nancy, France2Heinrich Heine Universtitat, Dusseldorf, Germany

October 9–10, 2019, Gothenburg

S. Pogodalla (with L. Kallmeyer and R. Osswald) A (Hybrid) Logical Approach to Frame Semantics Oct. 9–10, 2019 1 / 30

Outline

1 Frame SemanticsThe Frame HypothesisFrames and Logic

2 Hybrid LogicReminder on Modal LogicHybrid Logic

3 Compositional Frame SemanticsQuantificationFrame Decomposition


Frame Semantics The Frame Hypothesis

Frame Semantics

Frame

A (cognitive or linguistic) structure that represents a situation.

A frame is a data-structure representing a stereotyped situation (Minsky 1974, p.1)

I thought of each case frame as characterizing a small abstract ‘scene’ or ‘situation’,so that to understand the semantic structure of the verb it was necessary tounderstand the properties of such schematized scenes (Fillmore 1982, p.115)

I propose that frames provide the fundamental representation of knowledge inhuman cognition (Barsalou 1992)



The Frame Hypothesis (Lobner 2014)

H1 The human cognitive system operates with a single general format ofrepresentations

H2 If the human cognitive system operates with one general format ofrepresentations, this format is essentially Barsalou’s frames

Representation as frames in the human mind of:§ lexical linguistic expressions and their meanings§ complex linguistic expressions and their meanings

Verb meaning frames go beyond “case frames”. For instance§ Aspectual characteristics of the situation§ Structured relations between semantic arguments

Decompositional approach to meaning (Osswald and Van Valin 2014)



Frame ExampleThe man walked into the house

l0

motionl1

man

path

walking

l2

house

agent

mover

pathmanner

endp at-regionpart-of

goal


Frame Semantics Frames and Logic

Formal Representation of FramesFrames as base-labelled feature structures with types and relations (Kallmeyer and Osswald 2013)

l0

motionl1

man

path

walking

l2

house

agent

mover

pathmanner


goal

»

—

—

—

—

—

—

—

—

—

—

—

—

—

—

—

–

motion

agent 1

”

manı

mover 1

goal 2

«

house

at-region w

ff

path

«

path

endp v

ff

manner”

walkingı

fi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

fl

part-of ( v , w )

Labelled attribute-value description (LAVD) language

0 : motion ^0 ¨ agent fi 1 ^0 ¨ agent “ 0 ¨ mover ^0 ¨ goal fi 2 ^0 ¨ path : path^0 ¨ manner : walking ^ 1 : man2 : house ^x 0 ¨ path ¨ endp, 2 ¨ at-regiony : part-of


Frame Semantics Frames and Logic

Base-Labelled Feature Structures and LAVD LanguageProperties:

A frame is represented as a base-labelled feature structure

It is specified using the LAVD language

It is the most general base-labelled feature structures that satisfy the given LAVDformula (minimal first-order model)

Suitable to frame decomposition and composition

Variable free

We take the semantic structures associated with the syntactic structures as genuinesemantic representations, not as some kind of yet to be interpreted logicalexpressions (Kallmeyer and Osswald 2013)

However:

No means for explicit quantification

Referential entities are treated as definites (reflected by the naming of nodes)

Similar to other “framish” formalisms, e.g., AMR, dependency structures. . .

Proposal

Frames are considered as relational models Ñ modal logic

Feature structures specification require the hybrid logic language extension

Use hybrid logic binders to quantify


Hybrid Logic Reminder on Modal Logic

Modal Logic (Blackburn 1993; Areces and ten Cate 2007)Formulas

Rel “ FuncY PropRel is a set of functional and non-functional relational symbols,

Prop a set of propositional variables

Nom a set of nominals (node names),

Svar a set of state variables

Stat “ NomY Svar.

Definition (Formulas)

The language of formulas Forms is defined as:

Forms ::“ J | p | φ | φ1 ^ φ2 | xRyφ | s | @sφ | Óx .φ | Eφ | Dx .φ

where p P Prop, φ, φ1, φ2 P Forms, R P Rel, s P Stat, x P Svar.

Moreover, we define:

φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J

| p | φ | φ1 ^ φ2 | xRyφ | s | @sφ | Óx .φ | Eφ | Dx .φ



φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J | p

| φ | φ1 ^ φ2 | xRyφ | s | @sφ | Óx .φ | Eφ | Dx .φ

where p P Prop

, φ, φ1, φ2 P Forms, R P Rel, s P Stat, x P Svar.


φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J | p | φ | φ1 ^ φ2

| xRyφ | s | @sφ | Óx .φ | Eφ | Dx .φ

where p P Prop, φ, φ1, φ2 P Forms

, R P Rel, s P Stat, x P Svar.


φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J | p | φ | φ1 ^ φ2 | xRyφ

| s | @sφ | Óx .φ | Eφ | Dx .φ

where p P Prop, φ, φ1, φ2 P Forms, R P Rel

, s P Stat, x P Svar.


φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ



Modal Logic (Blackburn 1993; Areces and ten Cate 2007)Models

Definition (Model)

A model M is a triple xM, pRMqRPRel,V y such that:

M is a non-empty set,

each RM is a binary relation on M,

V : Prop

Y Nom

ÝÑ ℘pMq is a valuation

such that if i P Nom then V piq is asingleton.

An assignment g is a mapping g : Svar ÝÑ M.g xm is an assignment that differs from g at most on x and such that g x

mpxq “ m.For s P Stat, we also define rssM,g to be the only m such that V psq “ tmu if s P Nomand rssM,g

“ gpsq if s P Svar.



Modal Logic (Blackburn 1993; Areces and ten Cate 2007)Satisfaction Relation

Definition (Satisfaction relation)

Let M be a model, w P M

, and g an assignment for M.

The satisfaction relation ( is defined as follows:

M,

g ,

w ( J

M,

g ,

w ( p iff w P V ppq for p P PropM,

g ,

w ( φ iff M,

g ,

w * φM,

g ,

w ( φ1 ^ φ2 iff M,

g ,

w ( φ1 and M,

g ,

w ( φ2

M,

g ,

w ( xRyφ iff there is a w 1P M such that RM

pw ,w 1q and M,

g ,

w 1( φ

M, g ,w ( s iff w “ rssM,g for s P StatM, g ,w ( @sφ iff M, g , rssM,g

( φ for s P StatM, g ,w (Óx .φ iff M, g x

w ,w ( φM, g ,w ( Eφ iff there is a w 1

P M such that M, g ,w 1( φ

M, g ,w ( Dx .φ iff there is a w 1P M such that M, g x

w 1 ,w ( φ








M,

g ,

w ( JM,

g ,

w ( p iff w P V ppq for p P Prop

M,

g ,

w ( φ iff M,

g ,

w * φM,

g ,

w ( φ1 ^ φ2 iff M,

g ,

w ( φ1 and M,

g ,

w ( φ2

M,

g ,


pw ,w 1q and M,

g ,

w 1( φ






w 1 ,w ( φ








M,

g ,

w ( JM,

g ,


g ,

w ( φ iff M,

g ,

w * φ

M,

g ,

w ( φ1 ^ φ2 iff M,

g ,

w ( φ1 and M,

g ,

w ( φ2

M,

g ,


pw ,w 1q and M,

g ,

w 1( φ






w 1 ,w ( φ








M,

g ,

w ( JM,

g ,


g ,

w ( φ iff M,

g ,

w * φM,

g ,

w ( φ1 ^ φ2 iff M,

g ,

w ( φ1 and M,

g ,

w ( φ2

M,

g ,


pw ,w 1q and M,

g ,

w 1( φ






w 1 ,w ( φ


Hybrid Logic Hybrid Logic






Stat “ NomY Svar.






φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J | p | φ | φ1 ^ φ2 | xRyφ

| s | @sφ | Óx .φ | Eφ | Dx .φ




φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ



Hybrid Logic (Blackburn 1993; Areces and ten Cate 2007)Formulas





Stat “ NomY Svar.



Forms ::“ J | p | φ | φ1 ^ φ2 | xRyφ

| s | @sφ | Óx .φ | Eφ | Dx .φ




φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J | p | φ | φ1 ^ φ2 | xRyφ | s

| @sφ | Óx .φ | Eφ | Dx .φ

where p P Prop, φ, φ1, φ2 P Forms, R P Rel, s P Stat

, x P Svar.


φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J | p | φ | φ1 ^ φ2 | xRyφ | s | @sφ

| Óx .φ | Eφ | Dx .φ

where p P Prop, φ, φ1, φ2 P Forms, R P Rel, s P Stat

, x P Svar.


φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J | p | φ | φ1 ^ φ2 | xRyφ | s | @sφ | Óx .φ

| Eφ | Dx .φ



φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.



Forms ::“ J | p | φ | φ1 ^ φ2 | xRyφ | s | @sφ | Óx .φ | Eφ

| Dx .φ



φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ








Stat “ NomY Svar.






φñ ψ ” φ_ ψ

rRsφ ” xRy φ

A

φ ” E φ



Modal Logic (Blackburn 1993; Areces and ten Cate 2007)Models

Definition (Model)




V : Prop

Y Nom

ÝÑ ℘pMq is a valuation







Hybrid Logic (Blackburn 1993; Areces and ten Cate 2007)Models

Definition (Model)




V : PropY Nom ÝÑ ℘pMq is a valuation








Definition (Model)




V : PropY Nom ÝÑ ℘pMq is a valuation such that if i P Nom then V piq is asingleton.







Definition (Model)





An assignment g is a mapping g : Svar ÝÑ M.

g xm is an assignment that differs from g at most on x and such that g x






Definition (Model)






mpxq “ m.

For s P Stat, we also define rssM,g to be the only m such that V psq “ tmu if s P Nomand rssM,g





Definition (Model)















M,

g ,

w ( JM,

g ,


g ,

w ( φ iff M,

g ,

w * φM,

g ,

w ( φ1 ^ φ2 iff M,

g ,

w ( φ1 and M,

g ,

w ( φ2

M,

g ,


pw ,w 1q and M,

g ,

w 1( φ






w 1 ,w ( φ



Hybrid Logic (Blackburn 1993; Areces and ten Cate 2007)Satisfaction Relation


Let M be a model, w P M, and g an assignment for M.The satisfaction relation ( is defined as follows:

M,

g ,

w ( JM,

g ,


g ,

w ( φ iff M,

g ,

w * φM,

g ,

w ( φ1 ^ φ2 iff M,

g ,

w ( φ1 and M,

g ,

w ( φ2

M,

g ,


pw ,w 1q and M,

g ,

w 1( φ






w 1 ,w ( φ






M, g ,w ( JM, g ,w ( p iff w P V ppq for p P PropM, g ,w ( φ iff M, g ,w * φM, g ,w ( φ1 ^ φ2 iff M, g ,w ( φ1 and M, g ,w ( φ2

M, g ,w ( xRyφ iff there is a w 1P M such that RM

pw ,w 1q and M, g ,w 1

( φ






w 1 ,w ( φ









( φM, g ,w ( s iff w “ rssM,g for s P Stat

M, g ,w ( @sφ iff M, g , rssM,g( φ for s P Stat

M, g ,w (Óx .φ iff M, g xw ,w ( φ

M, g ,w ( Eφ iff there is a w 1P M such that M, g ,w 1

( φM, g ,w ( Dx .φ iff there is a w 1

P M such that M, g xw 1 ,w ( φ









( φM, g ,w ( s iff w “ rssM,g for s P StatM, g ,w ( @sφ iff M, g , rssM,g

( φ for s P Stat

M, g ,w (Óx .φ iff M, g xw ,w ( φ














w ,w ( φ

















w 1 ,w ( φ



Example

l0

motionl1

man

path

walking

l2

house

agent

mover

pathmanner


goal

l0 ^motion ^ xagentypl1 ^manq ^ xmoveryl1 ^ xgoalypl2 ^ houseq^

xmannerywalking ^ pDv w .xpathyppath^ xendpyvq^

@l2pxat-regionywq ^ @v pxpart-of ywqq


Compositional Frame Semantics Quantification

Quantification

agent

name

theme name

agent nametheme

agenttheme theme

kissing

man

John

womanMary

kissing man Peter

kissing woman

Sue

kissing

man

Paul

agent

name

name



Quantification

agent

name

theme name

agent nametheme

agenttheme theme

kissing

man

John

womanMary

kissing man Peter

kissing woman

Sue

kissing

man

Paul

agent

name

name

(1) a. John kisses Mary



Quantification

agent

name

theme name

agent nametheme

agenttheme theme

kissing

man

John

womanMary

kissing man Peter

kissing woman

Sue

kissing

man

Paul

agent

name

name

(1) a. John kisses Mary

b. Epkissing^ xagentypxnameyJohnq ^ xthemeypxnameyMaryqq



Quantification

agent

name

theme name

agent nametheme

agenttheme theme

kissing

man

John

womanMary

kissing man Peter

kissing woman

Sue

kissing

man

Paul

agent

name

name

(2) a. Every man kisses Mary



Quantification

agent

name

theme name

agent nametheme

agenttheme theme

kissing

man

John

womanMary

kissing man Peter

kissing woman

Sue

kissing

man

Paul

agent

name

name

(2) a. Every man kisses Mary

b.

A

pÓ i .man ñ Epkissing^ xagentyi ^ xthemeypxnameyMaryqqq



Quantification

agent

name

theme name

agent nametheme

agenttheme theme

kissing

man

John

womanMary

kissing man Peter

kissing woman

Sue

kissing

man

Paul

agent

name

name

(3) a. Every man kisses some woman



Quantification

agent

name

theme name

agent nametheme

agenttheme theme

kissing

man

John

womanMary

kissing man Peter

kissing woman

Sue

kissing

man

Paul

agent

name

name


b.

A

pÓ i .man ñ EpÓ i 1.woman^ Epkissing^ xagentyi ^ xthemeyi 1qqq



Quantification

agent

name

theme name

agent nametheme

agenttheme theme

kissing

man

John

womanMary

kissing man Peter

kissing woman

Sue

kissing

man

Paul

agent

name

name


b.

A


c. EpÓ i .woman^

A

pÓ i 1.man ñ Epkissing^ xagentyi 1^ xthemeyiqqq



Example

endp

path

mannerat-region

goal

agent

movername

pathmannermover

agent

goal

name

motion

path v1

walking

housew

manJohn

motion

v2

pathwalking

man Peter

part

-of part-of

endp



Example

endp

path

mannerat-region

goal

agent

movername

pathmannermover

agent

goal

name

motion

path v1

walking

housew

manJohn

motion

v2

pathwalking

man Peter

part

-of part-of

endp

(4) a. Every man walked to some house



Example

endp

path

mannerat-region

goal

agent

movername

pathmannermover

agent

goal

name

motion

path v1

walking

housew

manJohn

motion

v2

pathwalking

man Peter

part

-of part-of

endp


b.

A

pÓ i .man ñ p EpÓ i 1.house^ pDa g . Epmotion^ xagentya^ xmoverya^xgoalyg ^ xpathypath^ xmannerywalking^ @ai ^ pDr v w .event^xpathyppath^ xendpyvq ^ @r pxat-regionywq ^ @v pxpart-of ywq ^ @r pg ^ i 1

qqqqqqq



Example

endp

path

mannerat-region

goal

agent

movername

pathmannermover

agent

goal

name

motion

path v1

walking

housew

manJohn

motion

v2

pathwalking

man Peter

part

-of part-of

endp


b. EpÓ i 1.house^ p

A

pÓ i .man ñ pDa g . Epmotion^ xagentya^ xmoverya^xgoalyg ^ xpathypath^ xmannerywalking^ @ai ^ pDr v w .event^xpathyppath^ xendpyvq ^ @r pxat-regionywq ^ @v pxpart-of ywq ^ @r pg ^ i 1

qqqqqqq



Compositional Frame Semantics

Semantic types and constants

Types: e, s, t

Constants:

event, kissing,motion, John, . . . : t # : s Ñ txagenty, xthemey, xmovery, xpart-of y, . . . : t Ñ t @ : s Ñ t Ñ t^,ñ : t Ñ t Ñ t E,

A

: t Ñ t Ó, D : ps Ñ tq Ñ t

Lexical Semantics

JJohnK “ John JsomeK “ λP Q. EpÓ i .P ^ pQ p# iqqqJMaryK “ Mary JeveryK “ λP Q.

A

pÓ i .P ñ pQ p# iqqqJmanK “ man JkissesK “ λo s. Epkissing^ xagentys ^ xthemeyoqJwomanK “ woman





Types: e, s, t

Constants:

event, kissing,motion, John, . . . : t

# : s Ñ txagenty, xthemey, xmovery, xpart-of y, . . . : t Ñ t @ : s Ñ t Ñ t^,ñ : t Ñ t Ñ t E,

A


Lexical Semantics


A






Types: e, s, t

Constants:

event, kissing,motion, John, . . . : t # : s Ñ t

xagenty, xthemey, xmovery, xpart-of y, . . . : t Ñ t @ : s Ñ t Ñ t^,ñ : t Ñ t Ñ t E,

A


Lexical Semantics


A






Types: e, s, t

Constants:

event, kissing,motion, John, . . . : t # : s Ñ txagenty, xthemey, xmovery, xpart-of y, . . . : t Ñ t

@ : s Ñ t Ñ t^,ñ : t Ñ t Ñ t E,

A


Lexical Semantics


A






Types: e, s, t

Constants:

event, kissing,motion, John, . . . : t # : s Ñ txagenty, xthemey, xmovery, xpart-of y, . . . : t Ñ t @ : s Ñ t Ñ t

^,ñ : t Ñ t Ñ t E,

A


Lexical Semantics


A






Types: e, s, t

Constants:

event, kissing,motion, John, . . . : t # : s Ñ txagenty, xthemey, xmovery, xpart-of y, . . . : t Ñ t @ : s Ñ t Ñ t^,ñ : t Ñ t Ñ t

E,

A


Lexical Semantics


A






Types: e, s, t

Constants:


A

: t Ñ t

Ó, D : ps Ñ tq Ñ t

Lexical Semantics


A






Types: e, s, t

Constants:


A


Lexical Semantics


A






Types: e, s, t

Constants:


A


Lexical Semantics

JJohnK “ John

JsomeK “ λP Q. EpÓ i .P ^ pQ p# iqqq

JMaryK “ Mary

JeveryK “ λP Q.

A

pÓ i .P ñ pQ p# iqqq

JmanK “ man

JkissesK “ λo s. Epkissing^ xagentys ^ xthemeyoq

JwomanK “ woman





Types: e, s, t

Constants:


A


Lexical Semantics


A

pÓ i .P ñ pQ p# iqqqJmanK “ man

JkissesK “ λo s. Epkissing^ xagentys ^ xthemeyoq

JwomanK “ woman





Types: e, s, t

Constants:


A


Lexical Semantics


A






Types: e, s, t

Constants:


A


Lexical Semantics


A


JkissesK JMaryK JJohnK “Epkissing^ xagentypxnameyJohnq ^ xthemeypxnameyMaryqq





Types: e, s, t

Constants:


A


Lexical Semantics


A


pJeveryK JmanKq pλx .JkissesK JMaryK xq “

A

pÓ i .man ñ Epkissing^ xagentyi ^ xthemeypxnameyMaryqqq





Types: e, s, t

Constants:


A


Lexical Semantics


A


pJeveryK JmanKq pλx .pJsomeK JwomanKq pλy .JkissesK y xqq “

A






Types: e, s, t

Constants:


A


Lexical Semantics


A


pJsomeK JwomanKq pλy .pJeveryK JmanKq pλx .JkissesK y xqq “

EpÓ i .woman^

A

pÓ i 1.man ñ Epkissing^ xagentyi 1^ xthemeyiqqq



Compositional Frame Semantics (cont’d)

Lexical Semantics

walked “ λpp s.Da g . Epmotion^ xagentyp# aq ^ xmoveryp# aq^xgoalyp# gq ^ xpathypath^ xmannerywalking^ @as ^ ppp p# gqqq

to “ λn g .Dr v w .event^ xpathyppath^ xendpyp# vqq^@rxat-regionyp#wq ^ @vxpart-of yp#wq ^ @r pg ^ nq




Lexical Semantics



Jpevery manq pλx .psome houseqpλy .walked pto yq xqqK

“

A

pÓ i .man ñ p EpÓ i 1.house^ pDa g . Epmotion^ xagentya^ xmoverya^

xgoalyg ^ xpathypath^ xmannerywalking^ @ai ^ pDr v w .event^

xpathyppath^ xendpyvq ^ @r pxat-regionywq^

@v pxpart-of ywq ^ @r pg ^ i 1qqqqqqq




Lexical Semantics



Jpsome houseqpλy .pevery manq pλx .walked pto yq xqqK

“ EpÓ i 1.house^ p

A

pÓ i .man ñ pDa g . Epmotion^ xagentya^ xmoverya^

xgoalyg ^ xpathypath^ xmannerywalking^ @ai ^ pDr v w .event^

xpathyppath^ xendpyvq ^ @r pxat-regionywq^

@v pxpart-of ywq ^ @r pg ^ i 1qqqqqqq


Compositional Frame Semantics Frame Decomposition

Frame Decomposition and FrameNet (Osswald and Van Valin 2014)

Definitions of FN 1.5 frames of drying

Being dry: An [Item] is in a state of dryness (dehydrated, desiccated, dry, . . . )

Becoming dry: An [Entity] loses moisture with the outcome of being in a dry state(dehydrate, dry up, dry, . . . )

Cause to be dry: an [Agent] causes a [Dryee] (. . . ) to become dry (anhydrate, dehumidify,dehydrate, . . . )

«

Dry state

patient 2

ff

ÐÝ

»

—

–

Inchoation

result

«

Dry state

patient 2

ff

fi

ffi

fl

ÐÝ

»

—

—

—

—

—

—

—

—

—

–

Causation

cause

«

Activity

effector 1

ff

effect

»

—

–

Inchoation

result

«

Dry state

patient 2

ff

fi

ffi

fl

fi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

ffi

fl


Conclusion

Conclusion

Frame as a data-structure representing a situation

The semantic structure associated to a syntactic structure is a hybrid logicalformula that need to be further interpreted

Quantification does not belong to the frame language (contrary to Baldridge andKruijff 2002 or Kallmeyer and Richter 2014)

Inference

Link between dependency structures and logical representation (AMR,ECL—I. A. Mel’cuk, Clas, and Polguere 1995; I. Mel’cuk and Polguere 2018—, etc.)

Not variable free. . . But discourse referents are now made available!


Conclusion

Bibliographie I

Areces, Carlos and Balder ten Cate (2007). “Hybrid logics”. In: Handbook ofModal Logic. Ed. by Patrick Blackburn, Johan Van Benthem, and Frank Wolter.Vol. 3. Studies in Logic and Practical Reasoning. Elsevier. Chap. 14, pp. 821–868.doi: 10.1016/S1570-2464(07)80017-6.

Baldridge, Jason and Geert-Jan Kruijff (2002). “Coupling CCG and Hybrid LogicDependency Semantics”. In: Proceedings of 40th Annual Meeting of theAssociation for Computational Linguistics (ACL 2002). Philadelphia, Pennsylvania,USA: Association for Computational Linguistics, pp. 319–326. doi:10.3115/1073083.1073137. ACL anthology: P02-1041.

Barsalou, Lawrence (1992). “Frames, concepts, and conceptual fields”. In:Frames, fields, and contrasts: New essays in semantic and lexical organization.Ed. by Adrienne Lehrer and Eva Feder Kittey. Hillsdale: Lawrence ErlbaumAssociates, pp. 21–74.

Blackburn, Patrick (1993). “Modal Logic and Attribute Value Structures”. In:Diamonds and Defaults. Ed. by Maarten de Rijke. Vol. 229. Synthese Library.Springer Netherlands, pp. 19–65. isbn: 978-90-481-4286-6. doi:10.1007/978-94-015-8242-1_2.

Fillmore, Charles J. (1982). “Frame semantics”. In: Linguistics in the MorningCalm. Seoul, South Korea: Hanshin Publishing Co., pp. 111–137.


https://doi.org/10.1016/S1570-2464(07)80017-6

https://doi.org/10.3115/1073083.1073137

http://aclweb.org/anthology/P02-1041

https://doi.org/10.1007/978-94-015-8242-1_2

Conclusion

Bibliographie II

Gamerschlag, Thomas et al., eds. (2014). Frames and Concept Types. Vol. 94.Studies in Linguistics and Philosophy. Springer International Publishing. doi:10.1007/978-3-319-01541-5.

Kallmeyer, Laura and Rainer Osswald (2013). “Syntax-Driven Semantic FrameComposition in Lexicalized Tree Adjoining Grammars”. In: Journal of LanguageModelling 1.2, pp. 267–330. doi: 10.15398/jlm.v1i2.61.

Kallmeyer, Laura and Frank Richter (2014). “Quantifiers in Frame Semantics”. In:Formal Grammar. Ed. by Glyn Morrill et al. Vol. 8612. Lecture Notes in ComputerScience. Springer Berlin Heidelberg, pp. 69–85. doi:10.1007/978-3-662-44121-3_5.

Lobner, Sebastian (2014). “Evidence for frames from human language”. In:Frames and Concept Types. Applications in Language and Philosophy. Ed. byThomas Gamerschlag et al. Vol. 94. Studies in Linguistics and Philosophy 94.Dordrecht: Springer International Publishing. Chap. 2, pp. 23–67. isbn:978-3-319-01540-8. doi: 10.1007/978-3-319-01541-5_2.

Mel’cuk, Igor A., Andre Clas, and Alain Polguere (1995). Introduction a la

lexicologie explicative et combinatoire. Louvain-la-Neuve: Editions Duculot.


https://doi.org/10.1007/978-3-319-01541-5

https://doi.org/10.15398/jlm.v1i2.61

https://doi.org/10.1007/978-3-662-44121-3_5

https://doi.org/10.1007/978-3-319-01541-5_2

Conclusion

Bibliographie III

Mel’cuk, Igor and Alain Polguere (2018). “Theory and Practice of LexicographicDefinition”. In: Journal of Cognitive Science 19.4, pp. 417–470. hal open archive:halshs-02089593. url:http://cogsci.snu.ac.kr/jcs/index.php/issues/?uid=263&mod=document.

Minsky, Marvin (1974). A Framework for Representing Knowledge. Tech. rep. 306.MIT, AI Laboratory Memo. url: http://hdl.handle.net/1721.1/6089.

Osswald, Rainer and Robert D. Van Valin Jr. (2014). “FrameNet, FrameStructure, and the Syntax-Semantics Interface”. In: Frames and Concept Types.Ed. by Thomas Gamerschlag et al. Vol. 94. Studies in Linguistics and Philosophy.Springer International Publishing. Chap. 6, pp. 125–156. isbn: 978-3-319-01540-8.doi: 10.1007/978-3-319-01541-5_6.


http://hal.inria.fr/halshs-02089593

http://cogsci.snu.ac.kr/jcs/index.php/issues/?uid=263&mod=document

http://hdl.handle.net/1721.1/6089

https://doi.org/10.1007/978-3-319-01541-5_6

A (Hybrid) Logical Approach to Frame...

Documents

Transcript of A (Hybrid) Logical Approach to Frame...