Embed Size (px)
Citation preview
Networking the Asian WordNet on
WordNet Management System (WNMS)
Virach SornlertlamvanichNational Electronics and Computer Technology Center (NECTEC), Thailand, andThai Computational Linguistics Laboratory (TCL), NICT, [email protected]
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Motivation
Need of a computational ontology Implementation
Quick start approach Reusability
Less language resource
Online collaborative environment Social networking
Multilingual development
Cross language web service Sharing
Interoperability
Evaluation
DisseminationThe 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Approaches
Asian WordNet Development
Translation approach
Use of the existing bilingual dictionaries
Synset assignment
KUI for collaborative editing
WNMS (WordNet Management System)
Distributed WordNet service
Service for cross language WordNet retrieval
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Synset Assignment (CS=4)
Example:L0: เป้ าหมาย
E0: aimE1: target
S0: purpose, intent, intention, aim, designS1: aim, object, objective, targetS2: aim
Accept the Synset that includes more than one English Equivalent with confidence score of 4.
L0
E0
S0
S1
E1
S2
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Synset Assignment (CS=3)
Example:L0: จ้ องL1: เพ้ งมอง
E0: stareE1: gaze
S0: stareS1: gaze, stare
Synonym
Accept the Synset that includes more than one English Equivalent from the synonym of the target language with confidence score of 3.
L0 E0
S0
S1
E1
S2
L1
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Synset Assignment (CS=2)
Example:L0: สูติแพทย้
E0: obstetrician
S0: obstetrician, accoucheur
Accept the only Synset that includes the English Equivalent with confidence score of 2.
L0 E0 S0
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Synset Assignment (CS=1)
Example:L0: ช้ อง
E0: holeE1: canal
S0: hole, hollow S1: hole, trap, cakehole, maw, yap, gapS2: canal, duct, epithelial duct, channel
Accept more than one Synset that includes each of the English Equivalent with confidence score of 1. L0
E0
S0
S1
E1
S2
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
Asian WordNet Development
GWN
AWN
ApplicationsDictionaryOntologyCL-SearchMTSummarizationIE/IR….
KUI
Lookup
Discussion
Addition
Correction
Voting
Translation
WN merged-WN
X-English
X-English
X-English
Thai-English
X-English
X-English
X-English
Indonesian-English
August 2-4, 2010
KUI for AWN
KUI (Knowledge Unifying Editor)
In the initial stage, KUI was developed for collaborative editing to review and complete the translation.
Advantage
Suitable for building a community
Disadvantages
Translation is for word translation rather than sense translation
Cannot show the relation between senses
System is not fully distributed
August 2-4, 2010The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
WNMS for AWN
WNMS (WordNet Management System)
Sense based translation rather word based translation
Show the relation between senses
System is fully distributed connected through a standard Open API
Collaborative editing tools based on KUI concept
August 2-4, 2010The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
WMS for BalkaNet, GWC2004
WordNet Exploitation through a Distributed Network of Servers, I. D. Koutsoubos, and et. al.
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
WNMS (WordNet Management System)
August 2-4, 2010
Participation (Translate)
Input a word to search
Input a translated word, and select degree of confidence
Input comment or memo if have
Delete
August 2-4, 2010The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
1
3
2
Participation (Vote)
Read the comment or memo
Votevote up vote down
August 2-4, 2010The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
2
1
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
Distributed WordNet Service
Distribute the WordNet service node
Service node can be locally maintained
Synset ID (or Synset Offset) is the key to link between nodes
August 2-4, 2010
Representation of Synset Translation
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Types of Services ‘sense’
Thai Sense (Get word translation by POS and SYNSET_OFFSET)Service URI : http://th.asianwordnet.org/services/sense/output/[callback]/pos/synset_offsetService Name : senseParameter :
pos = PartOfSpeech {n,v,r,s}, synset_offset is an English Princeton WordNet v.3.0 offset, represented in 8 digits
http://th.asianwordnet.org/services/sense/xml/n/02958343
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Types of Services ‘dictionary’
E-Dictionary (Get word translation by word entry)Service URI :
http://th.asianwordnet.org/services/dictionary/output/[callback]/type_of_dict/search_wordService Name : dictionaryParameter : type_of_dict = {en2th, th2en}, search_word is a word you want to search
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Types of Services
Auto complete (Get a list of words existing in WordNet by prefix auto completion)Service URI :
http://th.asianwordnet.org/services/autocomplete/output/[callback]/language/search_wordService Name : autocompleteParameter : language = {en,th}, search_word is a word you want to get autocomplete (Result:limit 50 records found)
WN-Browser (Browse WordNet and its semantic relations)Service URI :
http://th.asianwordnet.org/services/browse/output/[callback]/language/search_wordService Name : browseParameter : language = {en,th}, search_word is a word you want to get all semantic relations
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Visualization of AWN(http://www.asianwordnet.org/)
Asian WordNet Visualization of Asian
WordNet
Function Cross language visualization
3 modes of visualization
Progress (# of word) Thai 80098
Lao 72672
Japanese 66648
Korean 65483
Myanmar 26033
Indonesian 21584
Vietnamese 17767
Mongolian 2283
Bengali 1775
Sinhala 117
Collaboration TCL
ADD members
English->Japanese
Thai->English
Thai->Indonesian
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Guideline in WordNet Translation
Word entry must be translated into the appropriate WORD(s) by avoiding phrase and meaning explanation.
Words in a Synset must be interchangeable.
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Translational Issues
There are many cases that a gloss need to be expressed in a phrase or explanation, especially in the case of technical terms and scientific vocabulary.
Ex. ChaperonPOS NounSynset chaperon, chaperoneGloss one who accompanies and
supervises a young woman or gatherings of young people
Thai ผู้ตามควบคุมหญิงสาว
These concepts are not general for Thai language
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
Translational Issues (cont.)
A gloss can be expressed by two or more Thai words. These words have the core meaning but occur in different context. Should it be divided into more specific concept?Ex. Appear
POS Verb
Synset appear, come outGloss be issued or published; "Did your latest book
appear yet?"; "The new Woody Allen filmhasn’t come out yet”
Thai T1 = ตีพิมพ้; T2 = ออกฉาย
T1 occurs in the context of printed matter
T2 occurs in the context of film or movie
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
AWN: A Platform for Collaboration
August 2-4, 2010The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
http://www.asianwordnet.org
Current
August 2-4, 2010The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
AWN Partnership
Hindi
IITB (Indian Institute of Technology, Bombay), India
Pushpak Bhattacharya <[email protected]>
Indonesian
BPPT (BADAN PENGKAJIAN DAN PENERAPAN TEKNOLOGI), Indonesia
Hammam Riza <[email protected]>
Japanese
NICT (National Institute of Information and Communications Technology), Japan
Hitoshi Isahara <[email protected]>, Kou Kuroda <[email protected]>
Nanyang Technological University (NTU), Singapore
Francis Bond <[email protected]>
Lao
NAST (National Authority of Science and Technology), Lao PDR
Valaxay Dalaloy <[email protected]>
Mongolian
NUM (National University of Mongolia), Mongolia
Purev Jaimai <[email protected]>
August 2-4, 2010The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
AWN Partnership
Burmese
MCF (Myanmar Computer Federation), Myanmar
Myint Myint Than <[email protected]>
Nepali
MPP (Madan Puraskar Pustakalaya), Nepal
Laxmi Pd Khatiwada <[email protected]>
Sinhala
UCSC (University of Colombo School of Computing), Sri Lanka
Ruvan Weerasinghe <[email protected]>
Thai
NECTEC (National Electronics and Computer Technology Center), Thailand
TCL (Thai Computational Linguistics Laboratory), Thailand
Virach Sornlertlamvanich <[email protected]>
Vietnamese
VAST (Vietnamese Academy of Science and Technology), Vietnam
Luong Chi Mai <[email protected]>
August 2-4, 2010The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
Conclusion and Future Work
Asian WordNet Community
Language resource conversion and alignment
Language technology sharing
Collaborative development platform
AWN and language technology web service
Applications on digital heritage understanding etc.
AsianWordnethttp://www.asianwordnet.org/
Join us!
The 2nd International Workshop with Mentors on Databases, Web and Information Management for Young Researchers, Aoyama Gakuin University, Tokyo, Japan
August 2-4, 2010
![Automated WordNet Construction Using Word Embeddings · (WOLF) [4] Universal Wordnet [5] Extended Open Multilingual Wordnet [6] Synset Representation Synset Representation + Sense](https://img.dokumen.tips/doc/110x75/5f0fb0bf7e708231d44567ef/automated-wordnet-construction-using-word-embeddings-wolf-4-universal-wordnet.jpg)
![Arabic WordNet: Current State and Future Extensionsnlp/papers/rodriguez08.pdf · Spanish WordNet within the framework of EuroWordNet project ([13], [5]). Later, the Catalan WordNet](https://img.dokumen.tips/doc/110x75/5c66e6bd09d3f2e33b8cc5f6/arabic-wordnet-current-state-and-future-nlppapersrodriguez08pdf-spanish.jpg)
