Pr

1
Automatically Populating Acception Lexical
Database through Bilingual Dictionaries and
Conceptual Vectors
PAPILLON 2002 Mathieu Lafourcade LIRMM - France
2
Overwiew Objectives

• Lexical soup
• what ? Bilingual dic Conceptual vectors
• which heuristics ?
• for what ? linking decision and quality
  assessment

3
Conceptual vectorsvector space

• An idea
• Concept combination a vector
• Idea space
• vector space
• A concept
• an idea a vector V
• with augmentation V neighboorhood
• Meaning space
• vector space v

?
4
Conceptual vectors Thesaurus

• H thesaurus hierarchy K concepts
• Thesaurus Larousse 873 concepts
• V(Ci) lta1, , ai, , a873gt
• aj 1/ (2 Dum(H, i, j))

1/4
1
1/4
1/4
1/16
1/16
1/64
1/64
2
6
4
5
Conceptual vectors Concept c4peace
peace
conflict relations
hiérarchical relations
society
The world, manhood
6
Conceptual vectors Term peace
c4peace
7
Angular distance

• DA(x, y) angle (x, y)
• 0 ? DA(x, y) ? ?
• if 0 then x y colinear same idea
• if ?/2 then nothing in common
• if ? then DA(x, -x) with -x anti-idea of x

x
x
?
y
8
Angular distance

• DA(x, y) acos(sim(x,y))
• DA(x, y) acos(x.y/xy))
• DA(x, x) 0
• DA(x, y) DA(y, x)
• DA(x, y) DA(y, z) ? DA(x, z)
• DA(0, 0) 0 and DA(x, 0) ?/2 by definition
• DA(?x, ?y) DA(x, y) with ?? ? 0
• DA(?x, ?y) ? - DA(x, y) with ?? lt 0
• DA(xx, xy) DA(x, xy) ? DA(x, y)

9
Thematic distance

• Examples
• DA(tit, tit) 0
• DA(tit, passerine) 0.4
• DA(tit, bird) 0.7
• DA(tit, train) 1.14
• DA(tit, insect) 0.62
• tit insectivorous passerine bird

10
Conceptual vector base
v
v
v
v
v
v
English entry base
French entry base
fleuve
  
river
  
rivière
acc y
map
acc
carte.2
carte.1
card.1
acc x
Malay entry base
Japanese entry base
acc z
sungai
acc x

  
  
  
  
  
  
Acception base
11
Conceptual vector base
v
v
v
v
v
v
French entry base
v
fleuve
v
fleuve
v
rivière
v

rivière

acc x
v
carte.2
v
carte.2
v
carte.1
v
carte.1
acc y
acc x
empty
acc z
  
  
  
Acception base
12
English Vectorized monolingualdictionary
English-French Bilingualdictionary
Left over meaning
demand
Association
v
demand
v
v
v
v
Vector space
Glosses 2
Equivalents 2
Glosses 2
Equivalents 2
Glosses 2
Equivalents 2
Glosses 2
Equivalents 2
Associations between definitions, vectors,
glosses and equivalents
13
Source Language
Target Language
Acceptions
already existing link
created link
W-SL i
equiv W-TL

W-TL
equiv W-SL,

  
warning if not close
vectors are close
W-SL i
W-TL
v
W-SL,
W-TL
v
  
W-TL
W-SL,
v
  
W-TL
W-SL,
v
  
left over acceptions
W-TL


  
14
Source Language
Target Language
Acceptions
W-TL1
Equiv W-SL,
v
  
W-TL1
Equiv W-SL,
v
  
v
W-TL1
Equiv W-SL,
  
W-TL1
Equiv

W-SL i
W-TL1,W-TL2,
v
  
W-TL2
Equiv W-SL,
v
  
W-TL2
Equiv W-SL,
v
  
v
W-TL2
Equiv W-SL,
  
W-TL2

Equiv
  

15
  
equiv W-TL
W-SL i
  
createdacception
  
  
W-SL i
equiv W-TL
refinement links
  
W-SL i
equiv W-TL


  
W-SL i
equiv W-TL

closest vector
W-SL i
equiv W-TL


  
W-SL i
equiv W-TL

16
W-SL i
equiv W-TL
  


W-SL j
equiv W-TL


  
W-SL j
equiv W-TL

1
equiv W-TL
W-SL i

  

  
W-SL j
equiv W-TL


  
W-SL j
equiv W-TL

2
W-SL i
equiv W-TL

  

  
W-SL j
equiv W-TL


  
equiv W-TL
W-SL j

closest vector
17
Conclusion

• System in continuous learning
• Evolving results
• Hopefully converging
• Assisting and begin assisted by
• Vectorized lexical functions
• Human annotators
• Toward
• Community of lexical agents
• Lexical knowledge negotiation