This paper describes a syntax oriented spoken Japanese understanding system named "SPOJUS-SYNO. At first this system makes word based Hidden-Markov-Models (HMM) automatically by concatenating syllable-based (trained) HMMs. Then a word lattice is hypothesized by using a word spotting algorithm and word-based HMMs for an input utterance. In SPOJUS-SYNO, the time-synchronous left-to-right parsing algorithm is executed to find the best word sequence from the word lattice according to syntactic & semantic knowledge represented by a context free semantic grammar. This system was implemented in the "UNIX-QA" task with the vocabulary size of 521 words. Experimental result shows that the sentence recognition / understanding rate was about 80 / 87% for six male speakers.
Bibliographic reference. Nakagawa, Seiichi / Hirata, Yoshimitsu / Murase, Isao (1991): "The syntax-oriented spoken Japanese understanding system SPOJOS-SYNO II", In EUROSPEECH-1991, 463-466.