Computers hit the scene in the 1940s, but did not become commonplace until the development of the Graphical User Interface in the early 1980s. The next big step towards making computers universally useful was expected to be Voice User Interfaces, enabling users to interact with computers using spoken dialog and freeing them from their keyboards and displays. However Voice User Interfaces have had only limited success.
This talk explains where the problems and challenges lie and how UTEP's Interactive Systems Group is working to overcome them by developing: techniques for making computers capable of following the normal human conventions for turn-taking, ways to infer user intentions and feelings at the sub-second level from subtle non-verbal signals, and methods for modeling the human mind well enough to know in advance what the user is likely to say. The talk will also overview the methods used for knowledge discovery, systems building, and experimentation.