Abstract

As speech recognition technology improves, many software designers are being challenged to design speech user inter faces. For me, one of the most challenging issues is “How do users know what they can say?” If you think about it, a speech-only interface poses the same problems as a command-line interface. The functionality of applications is hidden, and the boundaries of what can and cannot be done are invisible. Graphical interfaces were invented largely to make hidden functionality visible to the user. In a speech-only environment, such as over a telephone, it is not possible to display menus, show options, or highlight buttons. Instead, other techniques must be used to guide users through a successful interaction.