Volume 1, Issue 6 - June 2001
   
 

Answers to Your Questions About VoiceXML

By Jonathan Engelsma

(Continued from Part 1)

Q: How is internationalization addressed in VoiceXML?

A: The VoiceXML 1.0 specification is an application of XML 1.0 and therefore supports the universal character set standard known as Unicode. In theory this allows markup text associated with grammars and synthesized speech to be coded for most spoken languages.

Moving forward, future versions of the language are expected to support the 'xml:lang' attribute of the XML core standard for the <vxml> tag, as well as for input/output tags such as <grammar> and <prompt>. In addition VoiceXML 2.0 implementations will be expected to conform to the W3C Speech Synthesis Markup Language and the XML form of the W3C Speech Recognition Grammar Format, both of which utilize the 'xml:lang' attribute.

The xml:lang attribute can be assigned string values according to the three schemes defined in RFC 1766. These schemes include either a combination of the language and country codes defined by the ISO 639 and ISO 3166 standards, IANA-defined registrations, or user-defined codes. To put it simply, this attribute informs the interpreter of which language the data associated with the tag is written in. Whether or not the underlying platform supports the languages VoiceXML applications specify is implementation dependent.

In terms of additional support for developing internationalized application, one should keep in mind that VoiceXML markup is typically generated by server side applications. The same internationalization/localization facilities developers have at their disposal in their web server development environment (resource bundles, etc.) can also be utilized when generating VoiceXML markup.

back to the top

 

Copyright © 2001 VoiceXML Forum. All rights reserved.
The VoiceXML Forum is a program of the
IEEE Industry Standards and Technology Organization (IEEE-ISTO).