DTMF grammar support for VoiceXML

When developing VoiceXML applications that use touch-tone input, you can use the standard built-in DTMF grammar capability that is defined in the W3C VoiceXML 2.0 specification.

The Avaya VoiceXML feature also provides limited support for XML-formatted DTMF grammars as defined in the W3C Speech Recognition Grammar Specification (SRGS).

The following example illustrates what can be used in an XML DTMF grammar:

<xml version="1.0"?>
<grammar mode="dtmf" root="DIGIT" version="1.0" xml:lang="en-US">
<rule id="DIGIT" scope="public">
<one-of>
<item>1<tag>?name='one';digit='1'</tag></item>
<item>2<tag>?name='two';digit='2'</tag></item>
<item>3<tag>?name='three';digit='3'</tag></item>
</one-of>
</rule>
</grammar>

Additional considerations for Avaya IR Release 1.2.1 or later related to DTMF grammar code:

• Supports the use of the tags and syntax shown in the example above for XML DTMF grammars.
• The VoiceXML feature does not support the use of Javascript in XML DTMF grammars.
• Scripting for XML DTMF grammars is not supported. Therefore, the tag element should be defined as shown in the above example.
  • With the example code of <item>1<tag>?name='one'</tag></item>, a DTMF entry of 1 sets name to value one
  • If the following tag element syntax is used: <item>1<tag>name='one'</tag></item>, a DTMF entry of 1 sets name to value name='one'

Note:
If an application already uses code that has the format of ?key=value , change it to ?key='value' For example, tag="?name=three" would change to tag="?name='three'"

All DTMF grammar processing is done in the VoiceXML interpreter (VXI), not in the speech recognition engine. Because of the separate processing, the speech recognition grammar is separate from the DTMF grammar. If you choose to use a DTMF grammar with VoiceXML, two grammars need to be written, one for speech recognition processing and one for DTMF.

© 2006 Avaya Inc. All Rights Reserved.