Motorola Introduces Voice Browser

Motorola, Inc.’s Internet and Connectivity Services Division (ICSD) recently introduced VoxMLTM Voice Markup Language, designed to extend the reach of the Internet and intranets by enabling access to online content from any telephone. The new programming language, which allows software developers and Internet content providers to write voice applications which allows users to access the Web by voice, is the first product roll out from Motorola ICSD, a division the company formed in August. The division was formed to focus on developing solutions that simplify communications, including speech-enabled services, web-based services, content services, in-vehicle services, and paging information services. "The VoxML platform is a leading example of applying Motorola technology to the Internet and communications markets," said Merle Gilmore, president of Motorola Communications Enterprise. "This initiative demonstrates our commitment to enabling new service businesses. VoxML provides the means for exciting consumer services - allowing people to access the Web using their voice." Applications using VoxML will allow people to access online information such as personal banking information, weather forecasts, sports scores, stock quotes and internal company information over phones or computers using natural voice commands. Building consumer offerings using VoxML will help accelerate the emerging speech-enabled services market, an industry expected to grow to $3.4 billion by 2003. "VoxML will revolutionize the way people access online information and Web content," said Maria Martinez, Motorola ISCD General Manager and Vice President. "Motorola is committed to developing an open voice application for services that will help consumers more easily access information and simplify their lives." Several content providers are already using VoxML to extend their services. A voice-activated flight status service is one example of a VoxML application that provides convenient access to timely information. A user calling in to access airline information can say "Is the ABC Airline’s flight from Washington, DC on time?" The airline’s VoxML application interprets the voice request and translates it to a Web request. The application locates and publishes the requested information in VoxML, which is then translated from text into speech and delivered back to the user. In the future, VoxML will enhance the user interfaces for many similar applications. Development
VoxML applications may interface with encapsulated dialogue components such as SpeechObjects from Nuance Communications. SpeechObjects are reusable speech components that incorporate the dialogue and vocabulary that may be used in speech applications. "VoxML adds enormous potential to the development of speech applications by increasing the reach of voice-enabled services," said Ronald Croen, president and CEO of Nuance Communications. "Now with the power of voice, access to vital Web-based information is within anyone’s reach." Users call a voice "browser" by dialing a regular telephone number from any telephone. The call is connected and the user is able to interact with their specified Internet and Intranet applications hosted on any Web server. The diagram below shows the relationship between the user on the telephone, the voice browser, and the Web server. The interaction path is transparent to the user, much like surfing the Web via the computer on a desk today. [IMGCAP(1)] The voice browser runs on behalf of the user. The browser interacts with the user over a voice connection via the telephone network and with a Web server. Speech recognition and speech synthesis resources are available for use by the caller. The user interacts via a high-speed Internet connection. Apart from using voice and audio for the user interface, the voice browser behaves much like other Web browsers. It fetches data over the Internet using the Web's standard URL addressing scheme, stores "cookies" on behalf of the user, and caches frequently accessed pages. A content provider’s application runs on an existing Web server. No new software packages or upgrades are required on the Web server. Even with sophisticated speech synthesis, it is not practical to read the Web pages developed for typical graphical browsing on the Web. Furthermore, the free-form INPUT elements of HTML forms do not align well with telephony-grade speech recognition. For these reasons, Motorola has developed a markup language specifically for voice applications on the Web. With the VoxML™ language, a developer can express multi-step dialogues that collect commands, input values, and audio samples from the user and play recorded audio and synthesized speech to the user. The VoxML™ language is based on the W3C eXtensible Markup Language (XML) standard. Motorola is working closely with other leading telecommunications and speech technology developers to create a broadly supported markup language for voice applications. For more information, visit this website: http://VoxML.mot.com.
SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues