Page 72 - profile2012.indd

Research Faculty 研究人員
副研究員
莊庭瑞 Tyng-Ruey Chuang
Associate Research Fellow
Ph.D., Computer Science, New York University
Tel: +886-2-2788-3799 ext. 1613 Fax: +886-2-2782-4814
Email: trc@iis.sinica.edu.tw
http://www.iis.sinica.edu.tw/~trc/
攝影：張欽隆 Photo: Chin-Lung Chang
● Fellow, Berkman Center for Internet and Society, Harvard Univer- ● Assistant Research Fellow, Institute of Information Science, Aca-
sity, USA (2011/09 -- 2012/06) demia Sinica, Taiwan (1994/1 -- 2001/9).
● Associate Professor, Department of Information Management, ● Guest Researcher, Department of Computing Science, Chalmers
National Taiwan University, Taiwan (2008/02 -- ) University of Technology, Sweden (1993/9 -- 1993/12).
● PhD in Computer Science, New York University (1993).
● Associate Research Fellow, Institute of Information Science, ● MS in Computer Science, New York University (1990).
Academia Sinica, Taiwan (2001/10 -- ). ● BS in Computer Science and Information Engineering, National
Taiwan University (1984).
Research Description Publications
We do research in programming languages, Web technol- 1. Chin-Lung Chang, Yi-Hong Chang, Tyng-Ruey Chuang,
ogies, geo-spatial informatics, and Internet and society. Dong-Po Deng, and Andrea Wei-Ching Huang. Narrative geo-
spatial knowledge in ethnographies: Representation and rea-
We proposed a parametric content model for XML DTDs soning. In 3 International Conference on GeoSpatial Seman-
rd
(Document Type De nitions) and construct, automatically, tics, pages 188-203. Mexico City, Mexico, December, 2009.
their validation procedures. The model provides a basis for 2. Dong-Po Deng, Tyng-Ruey Chuang, and Rob Lemmens. Con- Discovery and Data Mining,” once considered the bible of
typeful XML programming in ML (a type-safe program- ceptualization of place via spatial clustering and co-occur- data mining. In 1998, I was invited to contribute a research
ming language with high-order functions and parametric rence analysis. In 2009 International Workshop on Location
modules) and leads to a theory of modular XML transfor- Based Social Networks, pages 49-56. Seattle, Washington,
mations. On the more practical side, we used XML and re- USA, November, 2009.
lated technologies, such as SVG (Scalable Vector Graphics) 3. Chia-Hsin Huang, Tyng-Ruey Chuang, Dong-Po Deng, and
and GML (Geography Markup Language), to implement Hahn-Ming Lee. Building GML-native Web-based Geo-
novel Web-based geo-spatial information systems. We are graphic Information Systems. Computers & Geosciences,
now looking into research issues in the collaborative pro- 35(9):1802-1816, September 2009.
duction of geo-spatial and spatiotemporal datasets. 4. Tyng-Ruey Chuang and Max Schaefer. Status Report: Lay-
ered Streaming XML Processing with Modules. In 2007 ACM
We investigate issues of program correctness and soft- SIGPLAN Workshop on ML, pages 53-58. Freiburg, Germa-
ware safety from the perspective of modular software ny, October, 2007.
construction. Modern functional languages such as ML 5. Tyng-Ruey Chuang and Jan-Li Lin. On modular transforma-
have built-in language constructs that support modular tion of structural content. In 2004 ACM Symposium on Docu-
programming. However, for software safety properties ment Engineering, pages 201-210. Milwaukee, Wisconsin,
that go beyond type safety, the modularity of a software USA, October 2004.
system usually plays a less important role. We aim to de- 6. Min-Bin Chen, Tyng-Ruey Chuang, and Jan-Jan Wu. Efficient
velop principles and techniques to support module-level parallel implementations of near Delaunay triangulation with
veri cation for ML-like languages. We are especially inter- High Performance Fortran. Concurrency and Computation:
ested in the “proofs-as-programs” approach to program Practice and Experience, 16(12):1143-1159, October 2004.
correctness as embodied in proof assistants (such as Coq) 7. Rong-Guey Chang, Tyng-Ruey Chuang, and Jenq Kuen Lee.
and their program extraction mechanisms. Support and optimization for parallel sparse programs with ar-
ray intrinsics of Fortran 90. Parallel Computing 304(4):527-
550, April 2004.
8. Tyng-Ruey Chuang, Y. S. Kuo, and Chen-Min Wang. Non-
intrusive object introspection in C++. Software: Practice and
Experience, 32(2):191-207, February 2002.
9. Tyng-Ruey Chuang. Generic validation of structural content
with parametric modules. In 2001 International Conference
on Functional Programming, pages 98-101. Florence, Italy,
September 2001.
研究人員
72 Research Faculty
研究員
許鈞南 Chun-Nan Hsu
Research Fellow
Ph.D., Computer Science, University of Southern California
Tel: +886-2-2788-3799 ext. 1801 Fax: +886-2-2782-4814
Email: chunnan@iis.sinica.edu.tw
http://www.iis.sinica.edu.tw/pages/chunnan
● Research Fellow, IIS, Academia Sinica (2008/5– ) ● Assistant Research Fellow, IIS, Academia Sinica (1998/1–2002/1)
● Visiting Research Scientist, IBM Research-Almaden (2005/1– ) ● Assistant Professor, CSE, ASU, U.S. (1996/1–1998/1)
● Associate Research Fellow, IIS, Academia Sinica (2002/6–2008/5) ● Research assistant, (ISD) ISI, USC, U.S. (1993/1–1996/1)
● Adjunct Assistant Professor, CSIE, NCTU (1999/1–2000/1) ● Research assistant, EE, USC, U.S. (1992/1–1993/1)
Research Description Publications
My research interests include machine learning, data min- 1. Peng JY, Lin CC, Chen YJ, Kao LS, Liu YC, Chou CC,
ing, databases and bioinformatics. I gave a plenary pres-
Huang YH, Chang FR, Wu YC, Tsai YS, Hsu CN. “Automatic
entation in the rst International Conference on Knowl- Morphological Subtyping Reveals New Roles of Caspases
edge Discovery and Data Mining (KDD) in 1995, the only in Mitochondrial Dynamics.”PLoS Computational Biology,
graduate student among ten other senior scientists as ple- 7(10):e1002212, 2011.
2. Kuo CJ, Ling HT, Hsu CN. “Soft tagging of overlapping high
nary presenters in the conference. In 1997, I was invited to confidence gene mention variants for cross-species full-text
contribute a chapter in the book “Advances in Knowledge gene normalization.” BMC Bioinformatics, vol.12(Suppl
8):S6, 2011.
3. Lin YS, Lin CC, Tsai YS, Ku TC, Huang YH and Hsu CN.
article to “Data Mining and Knowledge Discovery,” the rst “A Spectral Graph Theoretic Approach to Quantification and
scienti c journal specialized in the subject. I co-founded Calibration of Collective Morphological Differences in Cell
DeepSpot Intelligent Systems Inc. in Taiwan in 2001 and Images,” Bioinformatics, 26(12):i29-i37. Proceedings of the
th
18 International Conference on Intelligent Systems for Mo-
won three major government sponsored industrial awards lecular Biology ISMB 2010
in 2003.
Since 2005, I have been the principal investigator of the 4. Kuo CJ, Ling HT, Lin KT and Hsu CN.“BIOADI: a machine
learning approach to identifying abbreviations and defini-
Advanced Bioinformatics Core, National Research Pro- tions in biological literature.” BMC Bioinformatics, 10(Suppl
gram in Genomic Medicine, Taiwan, and leading one of 15):S7. Supplement for selected articles from the InCoB 2009
conference.
the largest research e orts in computerized drug design
and discovery in Taiwan. In 2006, the rst drug candidate 5. Hsu CN, Huang HS, Chang YM and Lee YJ. “Periodic step-
due to the use of the software my team developed was size adaptation in second-order gradient descent for single-
pass on-line structured learning,” Machine Learning, vol. 77,
commercialized. In 2007, our teams achieved the best no. 2-3, p.195-224,December, 2009.
scores in the BioCreative 2 text mining challenge, an in- 6. Hsu CN, Chang YM, Huang HS and Lee YJ. “Periodic step-
ternational contest sponsored by CNIO (Spanish Center of size adaptation for single-pass on-line learning,” NIPS 22,
Cancer Research), MITRE Co., and NCBI (National Center for pages 763-771, December, 2009.
Biotechnology Information, National Institutes of Health, 7. Hsu et al. “Integrating high dimensional bi-directional pars-
USA). The purpose of the contest was to evaluate whether ing models for gene mention taggih International Conference
software systems can extract knowledge from a large col- ISMB 2008.
lection of bio-medical research text automatically. In 2008, 8. Hsu et al. “Detection of the inferred interaction network in
my paper was accepted for oral presentation in ISMB 2008, hepatocellular carcinoma from EHCO (encyclopedia of
one of the most prestigious conferences in the eld of bio- hepatocellular carcinoma genes online).”BMC Bioinformat-
informatics and computational biology. Currently, I have ics,8(66):1471-2105, 2007.
been working on applying arti cial intelligence to dis- 9. Lin CC, Tsai YS, Lin YS, Chiu TY, Hsiung CC, Lee MI,
cover innovative cancer treatments and working on a new Simpson JC and Hsu CN. “Boosting multiclass learning with
repeating codes and weak detectors for protein subcellular
machine learning algorithm that can incrementally learn
from in nite many training examples. localization”Bioinformatics, 23(24):3374-3381.2007.
10. Yuan HY, Chiou JJ, Tseng WH, Liu CH, Liu CK, Lin YJ, Wang
HH, Yao A., Chen YT and Hsu CN.”FASTSNP: an always up-
to-date and extendable service for SNP function analysis and
prioritization” Nucleic Acids Research, vol. 34, Web Server
Issue, pages W635 - W641, July 2006.
73