The World Wide Web Consortium (W3C) on Monday published VoiceXML 2.0 as a proposed recommendation, which is the penultimate stage in the body's standardisation process.
"This is the jewel in the crown of the W3C's speech interface framework," said Dave Raggett, the technical lead for the W3C's voice-browser activity, who is on assignment from Canon. "Because the companies worked so hard together and cooperated so strongly in refining these specifications and developing the test suite, we can be sure we have interoperable standards for interactive voice response applications."
Companies editing the VoiceXML 2.0 proposed recommendation include Hewlett-Packard, Motorola, ScanSoft, IBM and Tellme Networks.
VoiceXML is most commonly drawn on to use voice commands instead of keystrokes to automate responses to phone calls made to businesses. Uses include call routing and information gathering and retrieval. The specification can be used in conjunction with the Internet or independently of it.
VoiceXML 2.0 made its last advance, to candidate recommendation, in January 2003. The W3C attributed the longer-than-usual wait to the complexity and size of the specification.
The W3C also had to establish a patent advisory group (PAG), VoiceXML 2.0's second, to iron out intellectual property disputes over the technology. In accordance with its patent policy, finalised in May of last year, the W3C chartered the PAG to resolve patent claims by Rutgers University, Philips, and Avaya Communications. In June, the PAG concluded its work and said the recommendation should continue.
The W3C said the proposed-recommendation phase should last four weeks.
