Multilingual Lexicon

1
Multilingual Lexicon

• Ye Hong Hoe

2
Abstract   Multilingual lexicon is a lexicon that
consists of more than one language. It is
essential for machine translation.  In our
system, we will build a English-Malay lexicon
based on lexical resources available in
UTMK. These lexical resources are "Kamus Inggeris
Melayu Dewan", "Kamus Dewan", WordNet and Longman
Dictionary of Contemporary English. In order to
build our English-Malay lexicon, we are going to
extract valuable information from the lexical
resources, such as entry or head word, part of
speech, definition, examples, translation pairs
and semantic relations.
3
Introduction

• Lexicon
• knowledge of words in a language
• Multilingual Lexicon
• lexicon that consists of more than 1 language
• essential for machine translation

4
Objective

• build a multilingual lexicon based on lexical
  resources available in UTMK
• base system
• English Malay lexicon

5
Lexical Resources

• Kamus Inggeris Melayu Dewan (KIMD)
• Kamus Dewan (KD)
• WordNet
• a combined lexicon that consists of
• Longman Dictionary of Contemporary English
  (LDOCE)
• KIMD
• WordNet

6
Crossing the Resources
??
English - Malay
Malay
KIMD
KD
English
WordNet
7
Multilingual LexiconStructure Overview
English Word
Malay Word
--
Sense disambiguation
Definition
Definition
Example
Example
Translation memory
Example
8
Discussion

• Categorization of entry
• Inflectional form same entry
• e.g. run, runs, ran, running
• Derivational form different entry
• e.g. agree, agreement
• Selection of Malay equivalent form
• e.g. sing, v menyanyi, v sing,
  v nyanyi, n
• menyanyi and nyanyi can be considered as
  different entry
• sing, v menyanyi, v would be better choice

9
Future Work

• Multilingual lexical representation
• Target word sense disambiguation

10
Thank You