Automatic generation of speech interfaces for Web-based applications

This paper describes a framework for automatically providing speech access to Web-based corporate applications that are accessible via SMS, WAP and Web browsers and are authored using a subset of HTML. A small subset of voice-related extensions was defined that application developers can make use of if they choose to support the voice channel as well. The voice user interface of the resulting application is then automatically generated and rendered using VoiceXML. This speeds up the application development process and minimizes the need for speech experts in the development cycle.