Innovative Computing

Automatic grammatical analysis of English

- Syntactic parsing based on statistical grammar induction -

Abstract

We need to grasp the subject and predicate of sentences to understand English, Japanese and other languages. However, it is not easy for a computer to perform such grammatical analysis automatically. We have developed grammatical analysis software based on the statistical acquisition of grammatical patterns from English resources. We employed this approach for parsing and achieved the best performance on international benchmark data. In the future, our aim is to enable computers to understand language and thus support human communication activities. As the first step towards our goal, we expect the grammatical analyzer to be useful for achieving high-quality machine translation and automatic language correction systems.

Poster


Please click the thumbnail image to open the full-size PDF file.

Reference

  1. H. Shindo, Y. Miyao, A. Fujino, M. Nagata, “Bayesian Symbol-Refined Tree Substitution Grammars for Syntactic Parsing,” in Proc. of the ACL, 2012.
  2. H. Shindo, A. Fujino, M. Nagata, “Insertion Operator for Bayesian Tree Substitution Grammars,” in Proc. of the ACL, 2011.

Presentor

Hiroyuki Shindo
Hiroyuki Shindo
Innovative Communication Laboratory
Yasuhisa Yoshida
Yasuhisa Yoshida
Innovative Communication Laboratory
Jun Suzuki
Jun Suzuki
Innovative Communication Laboratory
Tsutomu Hirao
Tsutomu Hirao
Innovative Communication Laboratory
Masaaki Nagata
Masaaki Nagata
Innovative Communication Laboratory