Aucun%20titre%20de%20diapositive

1
Guessing Hierarchies and Symbolsfor Word
Meanings throughHyperonyms and Conceptual Vectors
Mathieu Lafourcade LIRMM - France http//www.lir
mm.fr/lafourcade
2
Overwiew Objectives

• lexical semantic representations
• conceptual vector model (cvm)
• autonomous learning by the system
• from a given  semantic space  (ontology)
• Constructing texonomies
• Hierarchical - findind hyperonyms
• Multiple inheritance - views
• ambiguity as noise
• towards self contained WSD annotations
•  I made a deposit at the bank 
• ?  I made a deposit at the bankltgmoneygt

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
2D view of  meaning space 
product
cat
5
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
6
Conceptual vectors Concept c4peace
peace
conflict relations
hierarchical relations
society
The world, manhood
7
Conceptual vectors Term peace
c4peace
8

exchange
profit
finance
9
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
10
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)

11
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
12
Some vector operations

• Addition ? Z X ? Y
• zi xi yi vector Z is normalized
• Term to term mult ? Z X ? Y
• zi (xi yi)1/2
• vector Z is not normalized
• Weak contextualization ? Z X ? (X ? Y)
  ?(X,Y)
• Z is X augmented by its mutual information with
  Y

13
2D view of weak contextualization
X
Y
14
Autonomous learning 1/2

• set of known words K, set of unknow words U
• revise a word w of K OR (try to) learn a word w
  of U
• From the web for w ask for a def D
• specific sites dicts, synonyms list, etc. ? def
  analysis
• general sites google, etc. ? corpus analysis
• for each word wd of D
• if not in K then add wd to U AND add VO to V
• otherwise get the vector of wd AND add V(wd) to
  V
• compute the new vector of w from def(D) and V

98870 words for 400000 senses (vectors) learned
in 3 years
French
 ever  looping process
15
Autonomous learning 2/2
V
TXT
V
PH
V
N, GOV
V

ADJ,
V
insectivorous passerine bird
16
(No Transcript)
17
Hyperonyms identifications

• Extraction
• Try all terms
• too costly and unproductive
• Extract potential candidates
• From definitions, cooccurence lists etc.
• Ex Cand(emerald) precious stone, stone, beryl,
  gem,
• Evaluation of cand (m) to meaning (m)
• Contextualize ?(c,m) c ? (c ? m)
• Retain c such as ?(c,m) is the closest to m
• Loop extracting hyper helps identifying meanings

18
Pierre précieuse
Gemme/pierre précieuse
Gemme/bourgeon
Gemme/résine
v
v
v

béryl
closest vector
Émeraude/pierre précieuse
Émeraude/béryl
Émeraude/gemme
v
v
v
Pierre précieuse
Gemme/pierre précieuse
Gemme/bourgeon
Gemme/résine
v
v
v

béryl
Émeraude/pierre précieuse
Émeraude/béryl
Émeraude/gemme
v
v
v
19
Pierre précieuse
0.9
Gemme/pierre précieuse
Gemme/bourgeon
Gemme/résine
0.81

béryl
0.7
0.85
Émeraude/pierre précieuse
Émeraude/béryl
Émeraude/vert
Émeraude/couleur
Pierre précieuse

Couleur/matière
Couleur/sensation
0.9
Gemme/pierre précieuse

Vert/couleur des signaux
Vert/couleur
0.81
béryl
0.85
Émeraude/vert
Émeraude/béryl
20
Moyen de transport
artefact
hypo
aliment
animal
véhicule/Moyen de transport
véhicule/vecteur
nourriture
wagon
automobile
Cheval/moyen de transport
Viande/nourriture
Voiture/wagon
mammifère
hypo
Voiture/automobile
Cheval/viande
Cheval/mammifère
Cheval/unité de puissance
21
Last words

• Switching of representation
• From subsymbolic to symbolic
• and vice-versa ? readabily of symbols of
  words
• global and local test functions
• for vector quality assessment
• decision taking about number of meanings or
  views
• detectors when combined to lexical functions
  (antonymy, etc.)
• the basis for
• self adjustement toward a vector space of
  constant density
• wsd as a reduction of noise (in context or out of
  context)
• unification of ontologies
• self emergent structuration of terminology