"lexicon" Element

<lexicon>

Description

Sets the location of the user-defined pronunciation dictionary.

A lexicon file is an XML file following the PLS specification (http://www.w3.org/TR/pronunciation-lexicon/).

Syntax

Attributes

Attribute	Description
uri	URI of the dictionary. mandatory field
type	Media type of the dictionary. optional field
xml:lang	Locale. optional field
xml:id	May be used to give the lexicon reference an identifier that is unique to the document, allowing the element to be specified, via the <lookup> ref attribute as shown in the examples below. This is an SSML 1.1 extension.

Parent

<speak>

Children

None

Limitations/Restrictions

Note that the Lexicon element is not supported when using TTS2 voices.

Lexicons of types other than application/pls+xml are not supported. The Lexicon element must have a uri attribute specifying a URI that identifies the location of the pronunciation lexicon document. This location may utilize the optional xml:base attribute specified in the speak element (see examples below).

PLS Lexicon files are loaded separately for each synthesized SSML document. The only lexicons loaded are the ones declared by a <lexicon> element.

LumenVox Text-To-Speech supports PLS 1.0 lexicons referenced from SSML documents, as defined by Pronunciation Lexicon Specification (PLS) Version 1.0, W3C Recommendation 14 October 2008, http://www.w3.org/TR/2008/REC-pronunciation-lexicon-20081014/.

LumenVox supports both SSML 1.0 and an extension to the SSML 1.1 standard for supporting lexicon references as shown in the examples below. The corresponding SSML 1.1 <lookup> element is also supported (with corresponding ref attribute), also as shown below.

Example : SSML 1.0 style, local file reference to lexicon...

<?xml version="1.0"?>

<s>LumenVox has been recognized as a leading innovator of speech

technology since <say-as interpret-as="date"

format="year">2001</say-as>.</s>

</voice>

</speak>

Example : SSML 1.1 style, local file reference to lexicon, using lookup element...

<?xml version="1.0"?>

<s>LumenVox has been recognized as a leading innovator of speech

technology since <say-as interpret-as="date"

format="year">2001</say-as>.</s>

</voice>

</lookup>

</speak>

Example : SSML 1.1 style, xml:base specifying lexicon folder...

<?xml version="1.0"?>

<s>LumenVox has been recognized as a leading innovator of speech

technology since <say-as interpret-as="date"

format="year">2001</say-as>.</s>

</voice>

</lookup>

</speak>

Example : SSML 1.1 style, xml:base specifying lexicon folder on web server...

<?xml version="1.0"?>

<s>LumenVox has been recognized as a leading innovator of speech

technology since <say-as interpret-as="date"

format="year">2001</say-as>.</s>

</voice>

</lookup>

</speak>

Example : SSML 1.0 style, lexicon URI referencing web server (with xml:base)...

<?xml version="1.0"?>

<s>LumenVox has been recognized as a leading innovator of speech

technology since <say-as interpret-as="date"

format="year">2001</say-as>.</s>

</voice>

</speak>

Example : SSML 1.0 style, lexicon URI referencing web server (without xml:base)...

<?xml version="1.0"?>

<s>LumenVox has been recognized as a leading innovator of speech

technology since <say-as interpret-as="date"

format="year">2001</say-as>.</s>

</voice>

</speak>

Example Lexicon...

<?xml version="1.0" encoding="UTF-8" ?>
<lexicon version="1.0"
xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
alphabet="x-sampa" xml:lang="en-US">

<lexeme>
<grapheme>LumenVox</grapheme>
<phoneme>
lUUmEnvO:ks
</phoneme>
</lexeme>

<lexeme>
<grapheme>Cool Guys</grapheme>
<alias>
LumenVox
</alias>
</lexeme>

</lexicon>

You can copy this example lexicon as a starting point, but of course you will need to add your own lexemes to it. Please see our pages on the phonetic alphabets available for building the <phoneme> elements of the lexicon.

Multiple Lexicons in SSML 1.0

In order to work with multiple different lexicons in a single SSML 1.0 file, you need to separately group the portions of the file according to which lexicon should be used. For example:

<?xml version="1.0"?>
<speak version="1.0">


<lexicon uri="http://192.168.12.12/lexicons/first_lexicon.pls" />
<voice name="Amanda" xml:lang="en-US">
<s>Some phrase using words in first lexicon.</s>
</voice>


<lexicon uri="http://192.168.12.12/lexicons/second_lexicon.pls" />
<voice name="Amanda" xml:lang="en-US">
<s>Some phrase using words in second lexicon.</s>
</voice>


<lexicon uri="http://192.168.12.12/lexicons/third_lexicon.pls" />
<voice name="Amanda" xml:lang="en-US">
<s>Some phrase using words in third lexicon.</s>
</voice>

</speak>

As you can see in this example, sections of the files to be synthesized are contained in <voice> tags, which themselves are preceded by their corresponding lexicon references.

Note that the most recent lexicon reference is used for each <voice> section. It is also important to remember that <lexicon> tags must be children of <speak> elements (not within <voice> elements) and also <voice> elements are also children of <speak> elements, but should follow after any lexicon definition they use.

Note:

The syntax shown in this example is specific to SSML 1.0, since SSML 1.1 was changed to make use of the <lookup> element for selectively activating lexicons as described earlier in this article.

Was this article helpful?