In addition to VoiceXML 3.0, a working draft of which was introduced at the end of last year, the World Wide Web Consortium is working on a number of other specifications that should make it easier for developers to build and deploy more reliable and accessible speech platforms.
CCXML (Call Control XML) is designed to improve call handling at the network edge. The proposed specification will allow developers to better handle calls independently of the voice application.
CCXML complements VoiceXML by providing a way to declare how calls will be routed to applications and services within the network.
Using CCXML, developers will be able to build applications that allow features such as call conferencing and outbound calling.
In addition, CCXML will allow developers to perform call-handling tasks that VoiceXML doesnt provide a specification for, such as passing a call from one VoiceXML application to another and passing the call with messages, such as call-status events.
The W3C and the VoiceXML Forum, an industry consortium of speech and telephony vendors, are both working on specs that would provide VoiceXML with multimodal capabilities, allowing voice applications to be accessible through a Web browser.
The voice application framework used by Microsoft Speech Server 2004, SALT (Speech Application Language Tags) already has this capability.
The W3Cs EMMA (Extensible MultiModal Annotation) and the VoiceXML Forums XHTML+Voice markup language specifications will give developers a way to make VoiceXML applications more accessible to users with disabilities or via devices that make a Web interface impractical. VoiceXML Forum released Version 1.2 of the XHTML+Voice markup language last September; the W3Cs EMMA entered the working draft stage that same month.
The working draft specification for VoiceXML 3.0 includes a component, SSML (Speech Synthesis Markup Language), that provides a standard way to control some of the finer aspects of speech synthesis, including pronunciation, volume and rate of speech.