| The conjunction of several factors having occurred throughout the past
few years will make humans significantly change their behavior vis-à-vis machines.
In particular the use of speech technologies will become normal in
the professional domain, but also in everyday life. The performance of speech
recognition components has significantly improved: only within ten years we
have passed from systems able to recognize isolated words uttered by a single
speaker using a limited lexicon of around 50 words to systems able to recognize
continuous speech with an unlimited vocabulary uttered by any speaker;
or to systems able to carry a spontaneous dialog with a vocabulary of a few
thousands of words over the telephone, on a well-defined topic (e.g., information
on train or airplane schedules). The development of microelectronics,
considered to be at the origin of these significant results, also favors hardware
miniaturization enabling the integration of particularly time and memory
consuming signal processing algorithms into standard processors and portable
systems. Finally, the expanding computer and telephone networks allow the
general public an immediate location-independent access to large databases.
Under these circumstances, the state-of-the-art of research and achievements
in the domain of human-machine spoken language dialog systems, proposed
by Wolfgang Minker and Samir Bennacef, seems particularly relevant.
In fact, speech enables access to various information services from a simple
standard telephone: the number of calls (up to one and a half million) received
per month by systems developed at Philips, CSELT, Nuance, etc., shows the
adequacy of these vocal servers to the user needs. Speech, combined with a
touch screen, also improves the communication with information terminals,
rendering them more natural, effective and fast. |
