Voice Browser Packages Will Offer Hand-Off Access to Web Sites
From: PC Week - November 9, 1998 - page 36
By:   Herb Bethoney

Voice-enabled browsers offer the potential to deliver Internet information to
a much broader audience. By offering speech access to Web pages, voice
browsers will provide hands-free access to online information while driving a
car, for example. 

Web pages designed for viewing on a computer screen deliver a wealth of
information, but attempts to simply convert a Web page into speech - via a
text-to-speech application, for example - are laden with pitfalls. Elements
commonly found in electronic commerce Web sites, such as forms, frames,
diagrams and image maps, have no direct translation to speech and are not
easily navigated via the telephone. 

A voice browser accessible from a telephone must let users access Web links
and fill out form fields. Although IVR (interactive voice response) systems
have been around a long time, their menu-driven architecture doesn't lend
itself to the nonlinear aspects of surfing the Web. 

There are generally two approaches to bringing speech access to the Internet.
One approach is to extend HTML using style sheets. ACSS (Aural Cascading
Style Sheets), part of the World Wide Web Consortium's recommendation for the
CSS 2 specification, allows a document to be displayed aurally as well as
visually without requiring a separate Web page for each mode. ACSS is a
specification for reading Web pages to a user but doesn't provide a way for
developers to allow users to input speech. 

The other approach is to create a specific markup language for rendering
speech input as well as output on the Internet. This is the approach Motorola
Inc. has taken with its VoxML specification. 

But Motorola isn't the only player in the voice browser game. IBM has been
working with visually impaired computer users for many years to design screen
readers and provide accessibility to information technology. The result is
IBM Home Page Reader, a voice browser designed by Chieko Asakawa, a blind
researcher in IBM's Tokyo Research Labs. Home Page Reader, which is suitable
for voice input as well as output, was released in Japan in October 1997. IBM
Special Needs Systems, in Austin, Texas, is adapting it for North American
users and is adding support for HTML 4.0. 

Lucent Technologies Inc., of Murray Hill, NJ, is developing PhoneBrowser, a
speech recognition product for Internet service providers. PhoneBrowser is a
programmable platform that allows Web page authors to build IVR systems
without using expensive IVR equipment. 

PhoneBrowser reads Web pages to a caller via text-to-speech conversion. Users
control PhoneBrowser's voice browser by speaking over what the browser is
"saying," thus allowing a user to go to a specific point on a Web page
without having to wade through seemingly endless options. 

Siemens Corporate Research Inc.'s Liaison voice browser research effort is
aimed at providing drivers with access to Web-based information. Liaison is
an eyes-free and, for the most part, hands-free voice browser. Siemens, of
Princeton, NJ, is attempting to make listening to the Web like listening to
the radio, allowing drivers to make more productive use of their commuting
time. In an automobile, safety is the first concern, so Liaison uses a simple
voice navigation framework that demands minimal interaction.  

On the academic front, the Massachusetts Institute of Technology's Laboratory
for Computer Science has developed a speech system, called Jupiter, that
provides conversational access to weather information for 500-plus cities via
a standard telephone.