Taku Kudo
[English | Japanese]
Profile
- Name
- Taku Kudo
- Picture
- At Hokkaido
- Things Hooked on Recently
- Playing guitar and uklele
Research Topics
- Statistical Natural Language Processing
- Statistical Parsing, Statistical Dependency Analysis,
Chunking, Morphological Analysis
- Machine Learning (Support Vector Machines, Boosting, or
other Maximal Margin Classifiers)
- Data Mining, Text Mining
Publications
- Taku Kudo, Toshiyuki Hanaoka, Jun Mukai, Yusuke Tabata, and Hiroyuki Komatsu (2011)
Efficient dictionary and language model compression for input method editors, WTIM 2011
[PDF]
- Taku Kudo, Jun Suzuki, Hideki Isozaki (2005)
Boosting-based parse reranking with subtree features, ACL 2005
[PDF]
- Taku Kudo, Eisaku Maeda, Yuji Matsumoto (2004)
An Application of Boosting to Graph Classification, NIPS 2004
[PDF]
- Taku Kudo, Yuji Matsumoto (2004)
A Boosting Algorithm for Classification of Semi-Structured Text, EMNLP 2004 (to appear)
[PDF]
[PPT (slide)]
- Taku Kudo, Kaoru Yamamoto, Yuji Matsumoto (2004)
Appliying Conditional Random Fields to Japanese Morphological Analysis, EMNLP 2004 (to appear)
[PDF]
[PPT (slide)]
- Taku Kudo, Yuji Matsumoto (2004)
A Boosting Algorithm for Classification of Semi-Structured Text, EMNLP 2004 (to appear)
[PDF]
[PPT (slide)]
- Taku Kudo, Yuji Matsumoto (2003)
Fast Methods for Kernel-based Text Analysis, ACL 2003 (to appear)
[PDF]
[PPT (slide)]
- Taku Kudo, Yuji Matsumoto (2002)
Japanese Dependency Analyisis using Cascaded Chunking, CONLL 2002 in TAIPEI
[PDF]
[PPT (slide)]
- Taku kudo, Yuji Matsumoto (2001)
Chunking with Support Vector Machines, NAACL 2001
[PDF]
[PDF (slide)]
- Taku kudo, Yuji Matsumoto (2000)
Japanese Dependency Analysis
Based on Support Vector Machines, EMNLP/VLC 2000
[PDF]
[PDF (slide)]
- Taku kudo, Yuji Matsumoto (2000)
Use of Support Vector Learning
for Chunk Identification, CoNLL-2000
[Postscript]
Software
- TinySVM (SVMs package)
- YamCha (SVMs-based Chunker)
- CRF++ (CRF-based Chunker)
- CaboCha (Japanese Dependency Structure Analyzer (Japanese))
- Darts (Double Array pacakge (Japanese))
- FREQT (Frequent tree miner)
- BACT (A Boosting algorithm for classification of trees,
which is used in our paper on EMNLP2004-1.
- PrefixSpan (sequential pattern minier)
Miscellaneous
Links
- Machine Learning
- NLP
- Other Useful site
taku-spam@tahoo.org(for spam archive)