W3C to Expand Internationalization in Speech Synthesis Markup Language

The W3C released the results of the second Workshop on Speech Synthesis Markup Language, where speech experts from around the world presented ideas for expanding the range of languages supported by SSML 1.0.

 

The results include a new initiative to revise SSML 1.0 in ways that support a wider range of the world's languages, including the widely spoken languages of Mandarin, Hindi, Arabic, Russian, Hebrew, and other languages spoken in India and Asia.

These results reinforce important discoveries reached at the first SSML Workshop in Beijing late last year, which provided critical information on many Asian languages.

 

The announcement of the second workshop results serves as a call for participation to researchers around the world to join the effort to improve the specification.

 

An improved SSML will increase the ability of people worldwide to listen to synthesized speech through mobile phones, desktop computers, or other devices, extending the reach of computation and information delivery.

 

Expanding the Range of Languages Supported in Standards is Critical

The participants in the W3C Workshop reached conclusions that support the expansion of the SSML standard.

 

For example, the Workshop participants expressed the need to add to the standard the ability to represent features of spoken language, including tone, syllabic stress or accent, and duration in a machine-readable fashion. In some languages, these attributes are an important factor in determining meaning.

 

The goal of the next phase is to identify a few basic mechanisms that can extend the power of SSML to cover more of the world's languages.

 

W3C Invites Current and New Members to Join Efforts

W3C is moving forward on enhancing and expanding the capabilities of SSML, based on the results of the Workshop. Organizations, particularly those with native understanding of the languages of Japan, China, Korea, Russia and India are encouraged to join W3C and participate in the W3C Voice Browser Activity.