| |
Developing
X+V Applications Using the Multimodal Tools
Continued from previous
page...
Phase 1: Planning the application
It
is vital to plan the application in detail before the
actual development begins. During this phase, you make
high- and low-level design decisions that can assist
developers in creating standardized, well-behaved multimodal
applications, as well as reduce development time by
taking some of the guesswork out of interface design.
Planning will increase the usability of the interface,
reduce the end-user’s learning curve, and increase
the effectiveness of the application in meeting your
business needs. A thorough description of design decisions
is included in the VoiceXML Programmer's Guide, packaged
with the toolkit and the Voice Server SDK.
Just for starters, you will have to decide how and when
end-users will add information so that you can match
the visual and voice design. Keep in mind that all the
words users could say must be added to a grammar file
so that the speech recognition engine can match the
spoken words to usable data. Effective designs use simple
and consistent language for instructions, clear action
verbs, such as "Select your credit card type,"
and consistent phrasing throughout the application.
Designers can also investigate and take advantage of
the tools and features that are included in the toolkit.
One of the key features is the built-in set of Reusable
Dialog Components. When building an X+V application
using the toolkit, you can use dialog components, which
are ready-to-use sets of VoiceXML source code for common
functions. This will enable you to quickly and easily
add voice functions to your applications, thus saving
time and reducing the amount of code you need to write.
Dialog components can also be customized and reused
within an application or in other applications.
Phase 2: Creating the multimodal project and
file
With the design aspects in mind, you are now ready to
start an X+V application. After opening the toolkit,
you will use the New Project wizard to create a multimodal
project in which you will store all the files needed
for developing your application. Then you will use the
New File wizard to create your first multimodal file,
with file extension .mxml, the file format used by the
Multimodal Toolkit for an X+V application. The X+V editor
creates a new file in the toolkit with a pre-filled
heading and basic tags as shown in the following figure.
Phase 3: Adding the visual component
If you already have visual applications that you want
to voice-enable, you can use an import wizard, drag
& drop files into the Navigator, or cut & paste
existing HTML into the multimodal project. You can also
write XHTML code directly in the X+V editor.
All visual markup must comply with XHTML conventions.
This revision can be done before or after you add the
HTML in the toolkit. XHTML documents must be well formed,
with each element properly closed. All HTML elements
must be nested within the <html> root element.
See the XHTML 1.0 specification included
in the References section.
For best results, test the XHTML file to make sure it
is correctly revised before adding it into the project.
A good resource for testing and development is http://validator.w3.org.
You can use the Web site to check whether your page
is valid XHTML.
Phase 4: Adding the voice component
We will need to come back to the visual part later,
but now you are ready to begin coding the VoiceXML,
or voice portion. The VoiceXML code should be nested
within the <head> element of the XHTML document,
as you will see in the following example. You can write
the voice portion using two methods: coding VoiceXML
from scratch, or using the built-in Reusable Dialog
Components to add the voice input.
To demonstrate the options, we will use a simple example
of each to voice-enable the city field and the credit
card field.
Coding the easy way: Using Reusable Dialog Components
In place of, or in addition to, writing your own code,
you can use the built-in Reusable Dialog Components
to add common functions to your VoiceXML file. Each
subdialog includes sample calling code (file extension
.mxml) that you can copy and paste into your VoiceXML
file.
In the Reusable Dialog Components wizard (shown below),
you can select and customize a dialog component and
then import it within a VoiceXML <vxml:form> element,
with a unique id to reference in the HTML.
Continued...
back
to the top
Copyright
© 2001-2003 VoiceXML Forum. All rights reserved.
The VoiceXML Forum is a program of the
IEEE
Industry Standards and Technology Organization
(IEEE-ISTO).
|
