In this talk, Prof. Fujisaki will present a historical review for the concept of Spoken Language Processing (SLP), which was introduced by himself in 1986 in a Japanese National Project with the title “Advanced Human-Machine Interface through Spoken Language”. In spite of the world-wide acceptance of this concept, however, the author’s original idea seems to be less well understood, and most people still rely on the combination of speech signal processing and NLP in trying to solve their problems. Here Prof. Fujisaki will try to clarify the concept, and give his personal view on the future of the field.

In the second part, Prof. Fujisaki will discuss the problems that are inherent in the prevailing methods of automatic speech recognition (ASR). He will give a critical look at the widely accepted statistical models, indicate its shortcomings that are generally overlooked, and propose a new approach. He will particularly focus on the limitation of the word trigram model, which is a very primitive and insufficient way for modeling the actual process of message generation. Furthermore, he will challenge the task of the so-called ASR, and argue that speech understanding must precede speech transcription.