Page 77 - profile2012.indd

Research Faculty 研究人員

研究員 研究員
陳克健 Keh-Jiann Chen 陳孟彰 MengChang Chen

Research Fellow Research Fellow
Ph.D., Computer Science, State University of New York at Bu alo Ph.D., Computer Science, University of California, Los Angels

Tel: +886-2-2788-3799 ext. 1510 Fax: +886-2-2782-4814 Tel: +886-2-2788-3799 ext. 1802 Fax: +886-2-2782-4814
Email: kchen@iis.sinica.edu.tw Email: mcc@iis.sinica.edu.tw
http://www.iis.sinica.edu.tw/pages/kchen http://www.iis.sinica.edu.tw/pages/mcc/eindex.html

● Research Fellow, IIS, Academia Sinica (1989 - ) ● Editor-in-chief, The International Journal of Computational Linguistics ● Research Fellow (2007 - ) ● Associate Professor, NSYSU (1992-1993)
● Associate Research Fellow, IIS, Academia Sinica (1981/8 - 1989/7) and Chinese Language Processing (1996 - 2004) ● Deputy Director (1999-2002, 2008-2010) ● MTS, AT&T Bell Labs (1989-1992)
● Ph.D., Computer Science, State University of New York at Bu alo (1981) ● 2nd term president, The Computational Linguistic and Chinese Lan- ● Associate Research Fellow (1996-2007) ● Ph.D., Computer Science, UCLA (1989)

● Editor, International Journal of Computer Processing of Languages guage Processing Society (ROCLING) (1991 - 1993) ● Assistant Research Fellow (1993-1996)
Research Description Publications Research Description Publications

My research interests include Chinese language process- 1. Shu-Ling, Huang and Keh-Jiann Chen, 2011, “Semantic Anal- Currently, I am working on the following research projects. 1. “Querying Uncertain Minimum in Wireless Sensor Net-
ing, lexical semantics, lexical knowledge representation, ysis and Contextual Harmony of Durations”, Journal of Chi- (1) PLASH: A Platform for Location Aware Services with Hu- works”, Mao Ye, Ken Lee, Wang-Chien Lee, Xingjie Liu,
and corpus linguistics. We had been and continued in de- nese Linguistics. MengChang Chen, to appear on IEEE Transactions on Knowl-
veloping the research environments for Chinese natural 2. Chen, Wei-Te, Su-Chu Lin, Shu-Ling Huang, You-Shan man Computation. edge and Data Engineering.
language processing including CKIP Chinese lexical data- Chung, Keh-Jiann Chen, 2010,” E-HowNet and Automatic The PLASH platform is designed to help location-aware 2. “TSCAN: A Content Anatomy Approach to Temporal Topic
bases, Sinica corpus, Sinica Treebank, CKIP word segmen- Construction of a Lexical Ontology” Proceedings of Coling service providers deploy their applications easily so that Summarization”, Chien Chin Chen and MengChang Chen,
tation and pos tagging system, and Chinese parsers. The 2010. users can contribute their e orts and location-related data IEEE Transactions on Knowledge and Data Engineering,

current major research topic is compositional semantics. 3. Chung, You-shan and Keh-Jiann Chen, 2010, “Analysis of by using the services, which is the main di erence from pp170-183. January 2012.

In order to achieve automatic semantic composition for Chinese Morphemes and Its Application to Sense and Part-Of- traditional location-aware services. The collective e orts 3. “PLASH: A Platform for Location Aware Services with

Chinese language, we had developed a lexical semantic Speech Prediction for Chinese Compounds” Proceedings of Human Computation”, Yao Hua Ho, Yao Chuan Wu, and
representation system E-HowNet which is a frame-based ICCPOL conference 2010. and data generated by humans or sensors (or both) can be MengChang Chen, IEEE Communications Magazine, pp. 44-

entity-relation model to de ne lexical senses (concepts). used to help solve di cult location aware problems, such 51, December 2010.

The major features of E-HowNet are: 4. Chen,Keh-Jiann and Shu-Ling Huang, 2009, “A Step toward as real-time tra c estimation, environment sensing and
Compositional Semantics: E-HowNet a Lexical Semantic city pro ling. An LAS application deployed on PLASH can 4. “A novel approach for enhancing student reading comprehen-

a. Word senses (concepts) are de ned by not only Representation System” Proceedings of the 23 Pacifi c Asia exploit the functions provided by using the GUI provided sion and assisting teacher assessment of literacy”, Jun-Ming
rd

primitives but also any well-de ned concepts and Conference on Language, Information and Computation, pp. Chen, MengChang Chen and Yeali Sun, Computers and Edu-
conceptual relations. 1-8. by PLASH. In addition, application builders can contribute cation 55(3), pp. 1367-1382, November 2010.
b. Uniform sense representation for content words, 5. Bai, Ming-Hong, Jia-Ming You, Keh-Jiann Chen, and Jason their mature applications as a service component hosted 5. “An Adaptive Threshold Framework for Event Detection Us-
function words and phrases. S. Chang, 2009, “Acquiring Translation Equivalences of Mul- in system for other application builders to develop more ing HMM-based Life Profiles”. Chien Chin Chen, MengChang

c. Semantic relations are explicitly expressed. tiword Expressions by Normalized Correlation Frequencies” sophisticated applications. Chen and Ming-Syan Chen, ACM Transactions on Informa-
d. Semantic composition and decomposition capabili- Proceedings of EMNLP 2009. (2) Personalized Auto-Quiz for English Learning tion System, Feb. 2009.
ties. 6. Chia-Hung Tai, Jia-zen Fan, Shu-Ling, Huang and Keh-Jiann 6. “Using Incremental PLSI for Threshold Resilient Online
e. Near-canonical representations for lexical senses Chen, 2009, “Automatic Sense Derivation for Determinative- In this research, we study and implement a system to help Event Analysis”, Tzu-chuan Chou and MengChang Chen, pp.
and phrasal senses. Measure Compounds under the Framework of E-HowNet,” learner to comprehend their English capability. For a given 289-299, IEEE Transactions on Knowledge and Data Engi-
International Journal of Computational Linguistics and Chi- text, (which could be provided by learner or recommend- neering, March 2008.
The above features serve for the purpose of semantic com- nese Language Processing, vol. 14, #1, pp19-44.
position and natural language understanding. Currently ed by the system), the system give proper questionnaires 7. “An Incentive-based Fairness Mechanism for Multi-hop Wire-

we focus our attention on the problem of semantic deri- 7. Huang, Shu-Ling, and Keh-Jiann Chen, 2008, “Knowledge to examine the learner’s English capability and compre- less Backhaul Networks with Selfish Nodes”, Jeng Farn Lee,
vation of general compounds, since complex behaviors of Representation and Sense Disambiguation for Interrogatives hension on the text. Based on the learner’s scores on the Wanjiun Liao, and MengChang Chen, pp. 697-704, IEEE
semantic composition of morphemes can lead us to know in E-HowNet,“ International Journal of Computational Lin- quizzes, the system will adjust the level of learner which Transactions on Wireless Communications, February 2008..
better about general problem of semantic composition guistics & Chinese language Processing(IJCLCLP), Vol. 13, will be used in the forthcoming questionnaires genera- 8. ”WF2Q-M : Worst-case Fair Weighted Fair Queueing with
of constituents. To this end, we study morpho-syntactic #3, p. 255-278. tion. The related theoretical study includes text readability Maximum Rate Control”, Jeng Farn lee, MengChang Chen
behaviors of Chinese compounds by representing about 8. Shu-Ling Huang, Shih Yueh-Yin, Keh-Jiann Chen, 2008, evaluation, learner capability measurement, text under- and Yeali Sun, pp. 1403-1420, Volume 51, Issue 6, Computer
4,000 senses of morphemes (characters) in E-HowNet to “The Knowledge Representation for Comparison Words in standing, automatic English questionnaire generation, Networks, April 2007.
predict meanings and part-of-speeches of compounds by Extended-HowNet”, Language and Linguistics, Vol. 9, #2, P. and personalization scheme. 9. “A Framework of Handoffs in Wireless Overlay Networks
means of semantic composition rules. We hope that the 395-414. Based on Mobile IPv6”, Cheng Wei Lee, Leeming Chen,
ultimate goal of natural language understanding will be 9. Hsieh, Yu-Ming, Duen-Chi Yang, Keh-Jiann Chen, 2007,”Im- MengChang Chen and Yeali Sun, IEEE Journal on Selected
accomplished after future improvement and evolution of prove Parsing Performance by Self-Learning”, International Besides the above the projects, I have worked on some Areas in Communications, special issue on Wireless Overlay
the current E-HowNet. Journal of Computational Linguistics and Chinese Language Networks Based on Mobile IPv6, November 2005.
Processing, Vol. 12, #2, P. 195-216. projects, including wireless network, network security and
mobility management. 10. “ACIRD: Intelligent Internet Documents Organization and
Retrieval”, Shian-Hua Lin, MengChang Chen, Jan-Ming Ho,
and Yueh-Ming Huang, IEEE Transactions on Knowledge and
Data Engineering, p599-614, vol 14, no 3, 2002.
研究人員
76 Research Faculty
77