This project builds upon a long-term effort to create and distribute a comprehensive set of tools for research and development of spoken language systems. The effort has so far produced a large collection of speech corpora that are used in research, and has produced a portable software environment for authoring and using spoken language systems. The CARE project under consideration includes continuous envolutionary improvement of the systems and corpora, along with enhancements in some new directions. Three new lines of work that will be pursued are the establishment of evaluation criteria for evaluating new technologies in interactive systems, development of a new corpus of speech of young speakers, and development of training materials covering use of the toolkit in research and education.
