Parsing Japanese Spoken Sentences Based on HPSG

Kiyoshi Kogure


Abstract
An analysis method for Japanese spoken sentences based on HPSG has been developed. Any analysis module for the interpreting telephony task requires the following capabilities: (i) the module must be able to treat spoken-style sentences; and, (ii) the module must be able to take, as its input, lattice-like structures which include both correct and incorrect constituent candidates of a speech recognition module. To satisfy these requirements, an analysis method has been developed, which consists of a grammar designed for treating spoken-style Japanese sentences and a parser designed for taking as its input speech recognition output lattices. The analysis module based on this method is used as part of the NADINE (Natural Dialogue Interpretation Expert) system and the SL-TRANS (Spoken Language Translation) system.
Anthology ID:
W89-0214
Volume:
Proceedings of the First International Workshop on Parsing Technologies
Month:
August
Year:
1989
Address:
Pittsburgh, Pennsylvania, USA
Venues:
IWPT | WS
SIG:
SIGPARSE
Publisher:
Carnegy Mellon University
Note:
Pages:
132–141
Language:
URL:
https://www.aclweb.org/anthology/W89-0214
DOI:
Bib Export formats:
BibTeX MODS XML EndNote
PDF:
http://aclanthology.lst.uni-saarland.de/W89-0214.pdf