Third International Conference on Spoken Language Processing (ICSLP 94)

Yokohama, Japan
September 18-22, 1994

A Multi-Modal Dialogue System for Telephone Directory Assistance

Osamu Yoshioka, Yasuhiro Minami, Kiyohiro Shikano

NTT Human Interface Laboratories, Kanagawa, Japan

To evaluate our algorithm for large-vocabulary speaker-independent continuous speech recognition based on the HMM-LR technique and to collect dialogue data between computer and user, we have implemented a multi-modal dialogue system for telephone directory assistance. Our system uses three input devices (microphone, mouse, and keyboard) and two output devices (display and speaker). Twenty telephone number retrieval tasks were used to evaluate the system. In an evaluation involving twenty users, the task completion rate was 99.0%. About 75% of the users said they preferred this system to looking up telephone numbers in a telephone directory.

Full Paper

Bibliographic reference.  Yoshioka, Osamu / Minami, Yasuhiro / Shikano, Kiyohiro (1994): "A multi-modal dialogue system for telephone directory assistance", In ICSLP-1994, 887-890.