Acquisition of Semantic Classes for Adjectives from Distributional Evidence

1
Acquisition of Semantic Classes for Adjectives
from Distributional Evidence

• Gemma Boleda1, Toni Badia1, Eloi Batlle2
• GLiCom1, MTG2
• Universitat Pompeu Fabra
• Barcelona

2
general picture

• automatic classification of adjectives
• Catalan
• according to broad semantic characteristics
• clustering
• simple model part-of-speech n-grams

3
motivation

• Lexical Acquisition
• infer properties of words
• lexical bottleneck
• both symbolic and statistical approaches
• adjectives
• determining NP reference
• the French general
• establishing properties of entities
• this maimai is round and sweet

4
motivation

• initial motivation POS-tagging
• 55 remaining ambiguity involves adjectives
• general francès French general or general
  French one?
• observations
• general tendencies in syntactic behaviour of
  adjectives
• ... which correspond to broad semantic properties
• lets try a semantic classification
• low-level tasks (POS-tagging)
• initial schema for lexical semantic representation

5
approach

• no general, well established semantic
  classification
• have to test ours!
• clustering unsupervised technique
• groups objects according to feature distribution
• does not depend on pre-classification
• provides insight into the nature of the data
• shallow approach to syntax n-grams
• limited syntactic distribution (local
  relationship to arguments)
• gt test feasibility, limits

6
outline

• adjective syntax and semantic classification
• methodology
• experiment 1
• experiment 2
• conclusions and future work

7
outline

• adjective syntax and semantic classification
• methodology
• experiment 1
• experiment 2
• conclusions and future work

8
adjective syntax

• default function noun modifier (92)
• right of the noun (default position 72)
• some to the left (epithets 28)
• predicative uses unfrequent (7), but significant

9
two-way classification

• number of arguments
• unary red ball
• binary teacher jealous of Mary
• ontological kind (Ontological Semantics)
• basic red ball
• object pulmonary disease (gt lung)
• event constitutive property (gt constitutes)

10
rationale

• observation syntactic preferences correspond to
  semantic properties
• hypothesis we can use syntactic features to
  infer semantic classes

11
outline

• adjective syntax and semantic classification
• methodology
• experiment 1
• experiment 2
• conclusions and future work

12
data and procedure

• 2283 adjectives
• 50 times in 16 million word Catalan corpus
• lemma and morphological info
• cluster the whole set
• perform different tasks on different subsets
• tuning subset choose features
• Gold Standard evaluation and analysis

13
features and feature selection

• features
• empirically chosen from blind distribution
• double bigram, simplified POS-representation

• tuning subset 100 adjectives
• choose features (distribution)

14
Fig. A Feature selection
15
analysis

• Gold Standard
• 80 adjectives
• annotated by 3 human judges, acceptable agreement
  (92 and 84, .72 and .74 kappa)

16
outline

• adjective syntax and semantic classification
• methodology
• experiment 1
• experiment 2
• conclusions and future work

17
experiment 1 unary / binary

• final evaluation10 features, raw percentage
• clustering algorithm k-means (cosine)
• predictions (syntax-semantics)
• binary adjectives cooccur with prepositions more
  frequently than unary ones
• unary adjectives are more flexible

18
unary / binary results

• agreement with Gold Standard
• 97, kappa 0.87
• comparable to humans
• features

19
outline

• adjective syntax and semantic classification
• methodology
• experiment 1
• experiment 2
• conclusions and future work

20
experiment 2 basic / object / event

• final evaluation 32 features, normalisation
• clustering algorithm k-means (cosine)
• predictions (syntax-semantics)
• basic adjectives are flexible, appear further
  from the noun, work as epithets and occur
  predicative contexts
• object adjectives appear rigidly after the noun
• event adjectives tend to occur in predicative
  positions and do not act as epithets

21
basic / object / event results
object (yellow)

• agreement with Gold Standard
• 73, kappa 0.56
• lower than humans
• features

event (orange)
basic (red)
22
basic/object/event error analysis

• something has gone wrong!
• characterisation of event adjectives

binary!
basic adjectives with an object reading (polysemy)
unary event adjectives
binary event adjectives
23
outline

• adjective syntax and semantic classification
• methodology
• experiment 1
• experiment 2
• conclusions and future work

24
conclusions

• overall, results seem to back up
• use of syntax-semantics interface for adjectives
• linguistic predictions as to relevant features
  and differences across classes
• shallow modelling of syntactic distribution
• unary / binary piece of cake
• few binary adjectives, but worth spotting (denote
  relationships)

25
conclusions

• basic / object / event need reworking
• object adjectives seem to be the most robust
  class
• variation in basic adjectives (default class) gt
  subclassification?
• event adjectives seem to behave much like basic
  adjectives with respect to features chosen gt
  redefine class!

26
future work

• redefine classification
• redefine features in light of results
• integrate polysemy judgments into the experiment
  and analysis

27
Acquisition of Semantic Classes for Adjectives
from Distributional Evidence

• Gemma Boleda1, Toni Badia1, Eloi Batlle2
• GLiCom1, MTG2
• Universitat Pompeu Fabra
• Barcelona