This lecture will be a tutorial on using the SpeechBuilder toolkit to configure each of the language technology components necessary for building a basic mixed-iniative dialog system. The lecture will discuss motivations behind the SpeechBuilder toolkit and its role in the overall effort to the increase portability in building dialog systems. We will describe our recent efforts in enhancing SpeechBuilder that makes the creation and customization of a dialog system available at the web page. Much of this lecture will be tutorial-style on how to create a test systems using the graphical user interface in SpeechBuilder. This will prepare participants for the laboratory session in the afternoon.

Grace Chung is a research scientist at the Corporation for National Research Initiatives. Prior to this, Grace was a research assistant in the Spoken Language Systems Group at the MIT Laboratory for Computer Science. She received a Ph.D. in 2001, and a S.M. in 1997 from MIT, in Electrical Engineering and Computer Science. She has a B.Eng. in Electrical Engineering and a B.Sc. from the University of New South Wales, Sydney, Australia. Her research interests focus on many aspects of spoken dialog systems development. These include pronunciation, lexical and language modeling in speech recognition, as well as discourse and dialog modeling.

Joe Polifroni is a Research Scientist in the Spoken Language Systems Group. He is responsible for the various data collection efforts that the Spoken Language Systems Group has undertaken since 1989. In addition, he has been involved with all of the ARPA ATIS evaluations. He is also responsible for various system-related issues such as new accounts, mailing lists, and other general maintenance tasks. His interests include language generation, discourse and dialogue, and user interface issues.