Meshing Of XHTML, VoiceXML Is ProposedMeshing Of XHTML, VoiceXML Is Proposed

IBM, Motorola, and Web-browser company Opera have drafted a proposal for combining voiceXML, which allows navigation of Web pages through the use of voice commands, with xHTML, a mark-up language used to create Web pages.

The combination of both standards is what IBM refers to as a multimodal standard. One use of a multimodal application could be accessing data on an online travel site through a cell phone. Instead of typing in your departure and destination locations on the tiny keypad, you would speak your locations into the phone. The multimodal system would then reply to your request simultaneously through speech and by showing the list on the browser of your phone. "Multimodal gives you a visual output as well as voice output," says William Osborne, general manager of IBM voice systems.

The goal of a multimodal standard, Osborne says, is to help developers create mobile Internet applications using voiceXML and xHTML standards, which are already familiar to them.

IBM is not the only vendor pushing for ratification of voice standards. Cisco Systems, Microsoft, and other tech companies are part of an organization called SALT Forum, which is designed to standardize a new Web language that lets mobile device users access information online using voice input. SALT members plan to submit a language specification to the World Wide Web Consortium later this year. "We have not submitted a specification to the standards body yet, because we're trying to build as strong a submission as possible," says Rob Kassel a representative of the SALT Forum and a product manager with SpeechWorks International Inc., one of the founding members of the forum.

Osborne says IBM's multimodal proposal has been accepted by the W3C, and a multimodal working group is being formed within W3C to "develop a standard that everyone agrees with."