Page 59 - untitled
P. 59
д
Chen, Keh-Jiann Research Description Ӻᔊʧ
Research Description
Ӻᔊʧ
Construction of ontology and common sense ٝᗆܔ݊ɓঃࣛ൬ԫٙɽʈdҢࡁ йԸҬ̈ᇘႧจٙڌ༺˙όʿႧจٙΥϓዚ
knowledge databases is very time consuming. In the last ίཀ̘ɚɤεϋ೯࢝əʕ˖ஈଣਿᓾܔணމ͊ ՓfҢࡁɰਗ਼ҷഛԨΥɨ௰ࠠࠅٙɓԬٝ
ԸٙІਗʷٝᗆܔ͂ɨਿᓾfவԬਿᓾܔண ᗆݖν൚ၣeٝၣʿԫ࣪ݖၣd˸༺Ց༰
twenty some years, we have developed an infrastructure
̍ўᅺাႧࣘࢫe̩ഐዓ༟ࣘࢫe൚คࢫe ԳٙΥٝᗆڌ༺ӻ୕fҢࡁਗ਼೯࢝˸฿ׂމ
for Chinese language processing that includes part-of-
ʕ˖Ⴇجe൚คʱؓӻ୕ʿ̩࡚ؓኜഃfҢࡁ ʕːٙʕ˖ஈଣҦஔdה೯࢝ٙҦஔਗ਼л͜І
speech tagged corpora, treebank, Chinese lexical data-
ਗ਼л͜ҁϓٙਿᓾٝᗆၾҦஔԸІਗפ՟ၣ༩ ਗפ՟Ցٙ୕ࠇeႧԊႧجʿ੬ᗆৃࢹЪމ
bases, Chinese grammar, word identification systems,
˖ʕᒯўٙৃࢹdᓒ̂ତϞٝᗆݖԨܔͭ ਿᓾٝᗆ͜ʱؓ˖ٙ฿ׂഐԨᐝ༆˖
and sentence parsers. In the future we will utilize the Research Fellows
ჯਹٝᗆࢫʿ൚คٝᗆࢫfҢࡁਗ਼ஹഐʔΝٙ ٙจ່dආϾפ՟อٙٝᗆf˸ɪӉҖϓɓ
developed infrastructure to extract linguistic and do-
ٝᗆࢫҖϓɓࡈҁٙ฿ׂၣ˸৷ࠇၑዚપ ࡈІਗʷٙኪ୦ӻ୕dႧ˖ஈଣӻ୕̙͟І
main knowledge from various corpora and texts on the
ଣʿႧԊə༆ঐɢfίٝᗆڌ༺Ӻ˙ࠦҢࡁ ਗʱؓኪ୦อٝ˚һอٝᗆࢫdΝࣛɰᔟ͟
Web to enhance current knowledge databases. The tar-
ਗ਼ӺٝᗆݖٙਿᓾଣሞʿᇘႧจٙڌ༺ ٝᗆࢫٙһอᄣආəႧ˖ஈଣٙঐɢf
geted databases include general domain ontology, spe- ᅼό-ٝၣᓒ̂وf ᔟ͟ʱؓڐ່൚ٙฆࢨ
cial domain ontology, as well as lexical, syntactic, and
Research Fellows
ਿ ͉ ༟ ࣘ
ਿ ͉ ༟ ࣘ
semantic knowledge databases. The various databases
Selected Publications
will be inter-connected to form a ConceptNet for lan- Selected Publications
ᔖcc၈jӺࡰResearch Fellow (1989--)
guage processing and logical inference. For knowledge
௰৷ኪዝj Ph.D., CS, State University of New York representation we study the logical foundation of ontol- 1. Li, Shih-Min, Su-Chu Lin, and Keh-Jiann Chen. 2005 “Feature Rep- 13. Tsai Yu-Fang and Keh-Jiann Chen, 2003,” Context-rule Model for
POS Tagging”, Proceedings of PACLIC 17, pp146-151.
resentations and Logical Compatibility between Temporal Adverbs
at Buffalo (1981) ogy and fine-grain semantic representation. We also and Aspects”, International Journal of Computational Linguistics and 14. Keh-Jiann Chen, Jia-Ming You, Dee-Hwa Kao, & Cheng-Huei Wu,
Chinese Language Processing, Vol. 10, #4, P.445-458. 2003,” Design of Lexical Databases for Financial Domain”, Proceed-
ཥcc༑j+886-2-2788-3799 ext. 1510 study near-synonyms to identify fine-grain differences 2. Hsieh, Yu-Ming, Duen-Chi Yang, Keh-Jiann Chen, 2005, ”Linguis- ing of NLPKE 2003, PP 781-789.
tically-Motivated Grammar Extraction, Generalization, and Adapta- 15. Ma, Wei-Yun and Keh-Jiann Chen, 2003,” A Bottom-up Merging
ෂccॆj+886-2-2782-4814 between synonyms. These processes enable us to better tion”, Lecture Notes in Computer Science: Natural Language Pro- Algorithm for Chinese Unknown Word Extraction”, Proceedings of
ACL, Second SIGHAN Workshop on Chinese Language Processing,
cessing – IJCNLP 2005: Second International Joint Conference, Jeju
understand meaning representation and meaning com- Island, Korea, October 2005, Editors: Kam-Fai Wong, Jian Su, Oi Yee pp31-38.
ཥɿڦᇌjkchen@iis.sinica.edu.tw Kwong. P.177-187. 16. Ma, Wei-Yun and Keh-Jiann Chen, 2003,” Introduction to CKIP Chi-
position. We will remodel the current ontology struc- 3. Chen, K.J. & Yu-Ming Hsieh, 2005, ”Chinese Treebanks and Grammar nese Word Segmentation System for the First International Chinese
ၣccࠫjhttp://www.iis.sinica.edu.tw/pages/kchen tures of WordNet, HowNet, and FrameNet to achieve a Extraction”, Lecture Notes in Computer Science: Natural Language Word Segmentation Bakeoff”, Proceedings of ACL, Second SIGHAN
Processing – IJCNLP 2004: First International Joint Conference, Workshop on Chinese Language Processing, pp168-171.
better and more unified representation, called Extended- Hainan Island, China, March 22-24, 2004, Revised Selected Papers, 17. Chen, Keh-Jiann & Jia-Ming You, 2002, “A Study on Word Similarity
Editors: Keh-Yih Su, Jun’ichi Tsujii, Jong-Hyeok Lee, et al. P.655-661. Using Context Vector Models,” Computational Linguistics and Chi-
• Keh-Jiann Chen is a research fellow of the Institute HowNet. In addition, we will study modal logics and 4. Keh-Jiann Chen, Shu-Ling Huang, Yueh-Yin Shih, Yi-Jun Chen, nese Language Processing, Vol. 7, #2, pp. 37-58.
2005,” Extended-HowNet: A Representational Framework for 18. ಀᅆᙚeᄎ݇e৷eдd2002,© ˸൚ܛၾЧج
of Information Science, Academia Sinica and mem- integrate modal logic systems into a unified framework Concepts”, Proceedings of Ontolex 2005, pp.1-6. މ͉ٙʕ˖ਗ൚ІਗʱᗳӺ©, Computational Linguistics and
ber of the ACL, ACM, CLCS and the ROCLING and develop automated inference and theorem prov- 5. Shih, Yueh-Yin, Shu-Ling Huang, Yi-Jun Chen , Keh-Jiann Chen, Chinese Language Processing, Vol. 7, #1, pp.1-28.
society. He is the co-director of the Chinese Knowl- 2005,” Semantic Representation and Composition for Spatial Concepts 19. д, 2002,©ཥɿ൚Պၾ൚คٝᗆڌ༺©, ʕ˖ڦࢹኪజdVol.
ing methods based on the logical framework. We will in Extended-HowNet”, Proceeding of NLPKE 2005, pp.735-740. 16, #4,PP.1-11.
edge & Information Processing (CKIP) group at 6. Li, Shih-Min, Su-Chu Lin, Keh-Jiann Chen, 2005,”A Probe into 20. John Case, K. J. Chen, and Sanjay Jain, 2001 “Costs of General Pur-
focus on conceptual processing of Chinese documents. Ambiguities of Determinative-measure Compounds”, Proceedings of pose Learning”, Theoretical Computer Science, pages 455--473, Vol-
Academia Sinica. ROCLING 2005, pp.247-256. ume 259, Number 1—2.
The design of knowledge-based language processing 7. Tsai Yu-Fang and Keh-Jiann Chen, 2004,” Reliable and Cost-Effective 21. Chen, Keh-Jiann, Chu-Ren Huang, Feng-Yi Chen, Chi-Ching
• His current research interests include Chinese Pos-Tagging”, International Journal of Computational Linguistics & Luo,Ming-Chung Chang, Chao-Jan Chen, and Zhao-Ming Gao. 2001.
language processing, lexical semantics, and corpus systems will utilize statistical, linguistic, and common- Chinese Language Processing, Vol. 9 #1, pp83-96. Sinica Treebank: Design Criteria, Representational Issues and Imple-
linguistics. He is now developing research environ- sense knowledge provided by our evolving ConceptNet 8. Hua-jui Peng, Chi-ching Chen , Chiu-yu Tseng, and Keh-jiann Chen mentation. Anne Abeille (Ed.) Building and Using Syntactically An-
2004,”Predicting Prosodic Words from Lexical Words- a First Step notated Corpora. Language and Speech series. Dordrecht:Kluwer.
ments for Chinese natural language processing in- to parse the conceptual structures of sentences and Towards Predicting Prosody from Text” Proceedings of the 4 th Interna- 22. Huang, Chu-ren, K. Ahrens, Li-li Chang, Keh-Jiann Chen, M. C. Liu,
cluding Chinese lexical databases, Chinese corpora tional Symposium on Chinese Spoken Language Processing. Mei-Chih Tsai, 2000,” The Module-Attribute Representation of Verbal
interpret the meanings of sentences. Knowledge-based 9. You, Jia-Ming, Keh-Jiann Chen, 2004 “Automatic Semantic Role As- semantics: From Semantics to Argument Structure”’ International
and Chinese parsers. signment for a Tree Structure”, Proceedings of SIGHAN workshop. Journal of Computational Linguistics and Chinese Language Process-
language processing systems incorporate knowledge 10. Chu-Ren Huang, Chun-ling Chen, Cui-Xia Weng, and Keh-Jiann ing, Vol.5, #1, pp.19-46.
• Dr. Chen is one of the founding members of the Chen. 2004 “The Sinica Sense Management System: Design and 23. Chief, L.C., Chu-ren Huang, Keh-Jiann Chen, Mei-Chih Tsai, Li-
bases to form a learning system. Thus, language pro-
Computational Linguistic and Chinese Language Implementation” Proceedings of 5 th Lexical Semantics Workshop li Chang, 2000,” What Can Near Synonyms Tell Us”’ International
cessing systems increase their processing power due (CLSW-5), pp15-22. Journal of Computational Linguistics and Chinese Language Process-
Processing Society (ROCLING). He served as 2nd 11. Lin, Su-Chu, Shih-Min Li, and Keh-Jiann Chen. 2004 “Taxonomy of ing, Vol.5, #1, pp.47-60.
term president of the society from 1991 to 1993. to enhancement of the knowledge bases. Conversely, Fine-grain Semantic Roles for Nominal Modifi ers” Proceedings of 5 th 24. Li-li Chang, Keh-Jiann Chen, Chu-ren Huang, 2000,”Alternation
Lexical Semantics Workshop (CLSW-5) , pp259-263. Across Semantic Field: a Study of Mandarin Verb of Emotion”’ Inter-
He had been the editor-in-chief of the International the knowledge bases are evolving due to the automatic 12. Chen, Keh-Jiann, Chu-Ren Huang, Feng-Yi Chen, Chi-Ching national Journal of Computational Linguistics and Chinese Language
Journal of Computational Linguistics and Chinese Luo,Ming-Chung Chang, Chao-Jan Chen, and Zhao-Ming Gao, Processing, Vol.5, #1, pp. 61-80.
knowledge extraction made by language processing
Language Processing from 1996 to 2004, and is 2003, “Sinica Treebank: Design Criteria, Representational Issues and 25. ੵᘆᘆdдdර֢ʠd2000,©ဏႧਗ൚൚คႧจʱؓjڌ༺
systems. Implementation”. In Anne Abeille (Ed.) Treebanks Building and Us- ᅼόၾӺ˙ج© International Journal of Computational Linguis-
the editors of the Journal of Computer Processing ing Parsed Corpora. Language and Speech series. Dordrecht:Kluwer, tics and Chinese Language Processing, Vol.5, #1, pp. 1-18.
of Oriental Language and the Communications of pp231-248.
COLIPS.
48 49