Another abandoned server code base... this is kind of an ancestor of taskrambler.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 

3604 lines
149 KiB

<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta content=
"HTML Tidy for Linux/x86 (vers 1 September 2005), see www.w3.org"
name="generator" />
<meta http-equiv="Content-Type" content=
"text/html; charset=utf-8" />
<title>Pronunciation Lexicon Specification (PLS) Version
1.0</title>
<style type="text/css">
/*<![CDATA[*/
.diff-add { }
.diff-del { display: none }
.diff-chg { }
.diff-eval { background-color: #99FFFF; }
.diff-off { display: none }
.issues { font-weight: bold; color: green }
.ipa { font-family: "Lucida Sans Unicode", monospace; }
div.issue {
border-width: thin;
border-style: solid;
border-color: maroon;
background-color: #FFEECC;
color: maroon;
width: 95%; padding: 0.5em; }
div.issue h4 {
margin-top: 0;
font-weight: bold;
color: maroon;
}
pre {
font-family: monospace;
font-weight: bold;
white-space: pre;
background-color: rgb(204,204,255);
padding: 0.5em;
border: none;
margin-left: 0;
width: 90%;
}
pre.xml {
font-family: "Lucida Sans Unicode", monospace;
background-color: rgb(204,204,255);
border: solid black thin;
}
code {
color: green;
font-family: monospace;
font-weight: bold;
}
.RFC2119 {
text-transform: lowercase;
font-style: italic;
font-weight: bold;
}
/*]]>*/
</style>
<link rel="stylesheet" type="text/css" href=
"http://www.w3.org/StyleSheets/TR/W3C-REC" />
</head>
<body>
<div class="head">
<p><a href="http://www.w3.org/"><img height="48" alt="W3C" src=
"http://www.w3.org/Icons/w3c_home" width="72" /></a></p>
<h1 class="notoc" id="name">Pronunciation Lexicon Specification
(PLS) Version 1.0</h1>
<h2 class="notoc" id="date">W3C Recommendation 14 October 2008</h2>
<dl>
<dt>This version:</dt>
<dd><a href=
"http://www.w3.org/TR/2008/REC-pronunciation-lexicon-20081014/">
http://www.w3.org/TR/2008/REC-pronunciation-lexicon-20081014/</a></dd>
<dt>Latest version:</dt>
<dd><a href=
"http://www.w3.org/TR/pronunciation-lexicon/">http://www.w3.org/TR/pronunciation-lexicon/</a></dd>
<dt>Previous version:</dt>
<dd><a href=
"http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20080818/">
http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20080818/</a></dd>
<dt>Editor:</dt>
<dd>Paolo Baggia, Loquendo</dd>
<dt>Authors:</dt>
<dd>Paul Bagshaw, France Telecom</dd>
<dd>Daniel C. Burnett, Voxeo</dd>
<dd>Jerry Carter, Nuance</dd>
<dd>Frank Scahill, BT <i>(until 10 October 2001)</i></dd>
</dl>
<p>Please refer to the
<a href="http://www.w3.org/2008/10/pronunciation-lexicon-errata.html">
<strong>errata</strong></a>
for this document, which may include some normative
corrections.</p>
<p>See also
<a href="http://www.w3.org/2003/03/Translations/byTechnology?technology=pronunciation-lexicon">
<strong>translations</strong></a>.</p>
<p class="copyright"><a href=
"http://www.w3.org/Consortium/Legal/ipr-notice#Copyright">Copyright</a>
© 2008 <a href="http://www.w3.org/"><acronym title=
"World Wide Web Consortium">W3C</acronym></a><sup>®</sup>
(<a href="http://www.csail.mit.edu/"><acronym title=
"Massachusetts Institute of Technology">MIT</acronym></a>,
<a href="http://www.ercim.org/"><acronym title=
"European Research Consortium for Informatics and Mathematics">ERCIM</acronym></a>,
<a href="http://www.keio.ac.jp/">Keio</a>), All Rights
Reserved. W3C <a href=
"http://www.w3.org/Consortium/Legal/ipr-notice#Legal_Disclaimer">
liability</a>, <a href=
"http://www.w3.org/Consortium/Legal/ipr-notice#W3C_Trademarks">trademark</a>
and <a href=
"http://www.w3.org/Consortium/Legal/copyright-documents">document
use</a> rules apply.</p>
</div>
<hr title="Separator for header" />
<div>
<h2 class="notoc" id="abstract">Abstract</h2>
<p>This document defines the syntax for specifying <a href=
"#term-Pron-Lexicon">pronunciation lexicons</a> to be used by
<a href="#term-ASR">Automatic Speech Recognition</a> and
<a href="#term-TTS">Speech Synthesis</a> engines in voice
browser applications.</p>
</div>
<h2 id="Status">Status of this Document</h2>
<p><em>This section describes the status of this document at the
time of its publication. Other documents may supersede this
document. A list of current W3C publications and the latest
revision of this technical report can be found in the <a href=
"http://www.w3.org/TR/">W3C technical reports index</a> at
http://www.w3.org/TR/.</em></p>
<p>
This is the <a href=
"http://www.w3.org/2005/10/Process-20051014/tr.html#RecsW3C">Recommendation</a>
of
"Pronunciation Lexicon Specification (PLS) Version 1.0".
It has been produced by the
<a href="http://www.w3.org/Voice/">Voice Browser Working Group</a>,
which is part of the
<a href="http://www.w3.org/Voice/Activity.html">Voice Browser Activity</a>.
</p>
<p>Comments are welcome on <a
href="mailto:www-voice@w3.org">www-voice@w3.org</a> (<a
href="http://lists.w3.org/Archives/Public/www-voice/">archive</a>).
See <a href="http://www.w3.org/Mail/">W3C mailing list and archive
usage guidelines</a>.</p>
<p>The design of PLS 1.0 has been widely reviewed (see the
<a href="pls-disp.html">disposition of comments</a>)
and satisfies the Working Group's technical requirements.
A list of implementations is included in the
<a href="http://www.w3.org/Voice/2008/pls-ir/">
PLS 1.0 Implementation Report</a>,
along with the associated test suite.
The Working Group made a few editorial changes to the <a
href="http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20080818/">
18 August 2008 Proposed Recommendation</a> in response to comments.
Changes from the Proposed Recommendation can be found in
<a href="#AppD">Appendix D</a>.
</p>
<p><a href="#S2">Section 2. Pronunciation Alphabets</a> describes
the legal values of the <code>alphabet</code> attribute for
specifying a pronunciation alphabet. The Working Group is
requesting the creation of a Pronunciation Alphabet registry with
IANA so that pronunciation alphabets other than "ipa" can be also
used. The location of the registry will be provided at <a href=
"http://www.w3.org/2001/10/synthesis">http://www.w3.org/2001/10/synthesis</a>
when the registry becomes available. A future version of the PLS
specification may permit values from this registry to be used in
the <code>alphabet</code> attribute.</p>
<p>This document has been reviewed by W3C Members, by software
developers, and by other W3C groups and interested parties, and is
endorsed by the Director as a W3C Recommendation. It is a stable
document and may be used as reference material or cited from another
document. W3C's role in making the Recommendation is to draw
attention to the specification and to promote its widespread
deployment. This enhances the functionality and interoperability of
the Web.</p>
<p>This document was produced by a group operating under the <a href="http://www.w3.org/Consortium/Patent-Policy-20040205/" shape="rect">5 February 2004 W3C Patent Policy</a>. W3C maintains a <a rel="disclosure" href="http://www.w3.org/2004/01/pp-impl/34665/status" shape="rect">public list of any patent disclosures</a> made in connection with the deliverables of the group; that page also includes instructions for disclosing a patent. An individual who has actual knowledge of a patent which the individual believes contains <a href="http://www.w3.org/Consortium/Patent-Policy-20040205/#def-essential" shape="rect">Essential Claim(s)</a> must disclose the information in accordance with <a href="http://www.w3.org/Consortium/Patent-Policy-20040205/#sec-Disclosure" shape="rect">section 6 of the W3C Patent Policy</a>. </p>
<p>The sections in the main body of this document are normative
unless otherwise specified. The appendices in this document are
informative unless otherwise indicated explicitly.</p>
<hr />
<h2 id="contents">Table of Contents</h2>
<ul class='toc'>
<li class='tocline'>1. <a href="#S1">Introduction to
Pronunciation Lexicon Specification</a>
<ul class='toc'>
<li class='tocline'>1.1 <a href="#S1.1">How TTS Uses
PLS</a></li>
<li class='tocline'>1.2 <a href="#S1.2">How ASR Uses
PLS</a></li>
<li class='tocline'>1.3 <a href="#S1.3">How VoiceXML
Applications Use PLS</a></li>
<li class='tocline'>1.4 <a href="#S1.4">What PLS does not
Support</a></li>
<li class='tocline'>1.5 <a href="#S1.5">Glossary of
Terms</a></li>
</ul>
</li>
<li class='tocline'>2. <a href="#S2">Pronunciation
Alphabets</a></li>
<li class='tocline'>3. <a href="#S3">PLS Documents</a>
<ul class='toc'>
<li class='tocline'>3.1 <a href="#S3.1">Document
Form</a></li>
<li class='tocline'>3.2 <a href="#S3.2">Conformance</a>
<ul class='toc'>
<li class='tocline'>3.2.1 <a href="#S3.2.1">Conforming
Pronunciation Lexicon Specification Documents</a></li>
<li class='tocline'>3.2.2 <a href="#S3.2.2">Using PLS
with other Namespaces</a></li>
<li class='tocline'>3.2.3 <a href="#S3.2.3">Conforming
Pronunciation Lexicon Specification Processors</a></li>
</ul>
</li>
</ul>
</li>
<li class='tocline'>4. <a href="#S4">Pronunciation Lexicon
Markup Language Definition</a>
<ul class='toc'>
<li class='tocline'>4.1 <a href=
"#S4.1"><code>&lt;lexicon&gt;</code> Element</a></li>
<li class='tocline'>4.2 <a href=
"#S4.2"><code>&lt;meta&gt;</code> Element</a></li>
<li class='tocline'>4.3 <a href=
"#S4.3"><code>&lt;metadata&gt;</code> Element</a></li>
<li class='tocline'>4.4 <a href=
"#S4.4"><code>&lt;lexeme&gt;</code> Element</a></li>
<li class='tocline'>4.5 <a href=
"#S4.5"><code>&lt;grapheme&gt;</code> Element</a></li>
<li class='tocline'>4.6 <a href=
"#S4.6"><code>&lt;phoneme&gt;</code> Element</a></li>
<li class='tocline'>4.7 <a href=
"#S4.7"><code>&lt;alias&gt;</code> Element</a></li>
<li class='tocline'>4.8 <a href=
"#S4.8"><code>&lt;example&gt;</code> Element</a></li>
<li class='tocline'>4.9 <a href="#S4.9">Multiple
Pronunciations for ASR and TTS</a>
<ul class='toc'>
<li class='tocline'>4.9.1 <a href="#S4.9.1">Multiple
Pronunciations for ASR</a></li>
<li class='tocline'>4.9.2 <a href="#S4.9.2">Multiple
Pronunciations for TTS</a></li>
<li class='tocline'>4.9.3 <a href="#S4.9.3">Examples of
Multiple Pronunciations</a></li>
</ul>
</li>
</ul>
</li>
<li class='tocline'>5. <a href="#S5">Examples</a>
<ul class='toc'>
<li class='tocline'>5.1 <a href="#S5.1">Simple Use
Case</a></li>
<li class='tocline'>5.2 <a href="#S5.2">Multiple
Pronunciations</a></li>
<li class='tocline'>5.3 <a href="#S5.3">Multiple
Orthographies</a></li>
<li class='tocline'>5.4 <a href="#S5.4">Homophones</a></li>
<li class='tocline'>5.5 <a href="#S5.5">Homographs</a></li>
<li class='tocline'>5.6 <a href="#S5.6">Pronunciation by
Orthography (Acronyms, Abbreviations, etc.)</a></li>
</ul>
</li>
<li class='tocline'>6. <a href="#S6">References</a>
<ul class='toc'>
<li class='tocline'>6.1 <a href="#S6.1">Normative
References</a></li>
<li class='tocline'>6.2 <a href="#S6.2">Informative
References</a></li>
</ul>
</li>
<li class='tocline'>7. <a href="#S7">Contributors and
Acknowledgements</a></li>
<li class='tocline'>Appendix A. <a href="#AppA">Schema for
Pronunciation Lexicon Specification</a> (normative)</li>
<li class='tocline'>Appendix B. <a href="#AppB">MIME Type and
File Suffix</a> (normative)</li>
<li class='tocline'>Appendix C. <a href="#AppC">Issues in
Retrieving Lexical Content</a> (informative)</li>
<li class='tocline'>Appendix D. <a href="#AppD">Changes</a>
(informative)</li>
</ul>
<h2 id="S1">1. Introduction to Pronunciation Lexicon
Specification</h2>
<p><i>This section is informative.</i></p>
<p>The accurate specification of pronunciation is critical to the
success of speech applications. Most Automatic Speech Recognition
(<a href="#term-ASR">ASR</a>) and Text-To-Speech (<a href=
"#term-TTS">TTS</a>) engines internally provide extensive high
quality <a href="#term-Lexicon">lexicons</a> with pronunciation
information for many words or phrases. To ensure a maximum
coverage of the words or phrases used by an application,
application-specific pronunciations may be required. For example,
these may be needed for proper nouns such as surnames or business
names.</p>
<p>The Pronunciation Lexicon Specification (PLS) is designed to
enable interoperable specification of pronunciation information
for both <a href="#term-ASR">ASR</a> and <a href=
"#term-TTS">TTS</a> engines. The language is intended to be easy
to use by developers while supporting the accurate specification
of pronunciation information for international use.</p>
<p>The language allows one or more pronunciations for a word or
phrase to be specified using a standard pronunciation alphabet or
if necessary using vendor specific alphabets. Pronunciations are
grouped together into a PLS document which may be referenced from
other markup languages, such as the <a href="#term-SRGS">Speech
Recognition Grammar Specification</a> [<a href=
"#ref-SRGS">SRGS</a>] and the <a href="#term-SSML">Speech
Synthesis Markup Language</a> [<a href="#ref-SSML">SSML</a>].</p>
<p>In its most general sense, a <a href=
"#term-Lexicon">lexicon</a> is merely a list of words or phrases,
possibly containing information associated with and related to
the items in the list. This document uses the term <a href=
"#term-Lexicon">"lexicon"</a> in only one specific way, as
<a href="#term-Pron-Lexicon">"pronunciation lexicon"</a>. In this
particular document, "lexicon" means a mapping between words (or
short phrases), their written representations, and their
pronunciations suitable for use by an <a href="#term-ASR">ASR</a>
engine or a <a href="#term-TTS">TTS</a> engine. Pronunciation
lexicons are not only useful for voice browsers; they have also
proven effective mechanisms to support accessibility for persons
with disabilities as well as greater usability for all users.
They are used to good effect in screen readers and user agents
supporting multimodal interfaces.</p>
<h2 id="S1.1">1.1. How TTS Uses PLS</h2>
<p>A <a href="#term-TTS">TTS</a> engine aims to transform input
content (either text or markup, such as <a href=
"#term-SSML">SSML</a>) into speech. This activity involves
several processing steps:</p>
<ul>
<li>Text normalization</li>
<li>Word pronunciation (lexical stress, phonetic
transcription)</li>
<li>Sentence structure (intonation, rhythm)</li>
<li>Sentence level modification in phonetic transcription
(co-articulation)</li>
<li>Computation of prosodic parameters</li>
<li>Generation of the acoustic signal</li>
</ul>
<p><a href="#term-SSML">SSML</a> enables a user to control and
enhance <a href="#term-TTS">TTS</a> activity by acting through
<a href="#term-SSML">SSML</a> elements on these levels of
processing (see [<a href="#ref-SSML">SSML</a>] for details).</p>
<p>PLS is intended to be the standard format of the documents
referenced by the <a href=
"#S4.4"><code>&lt;lexicon&gt;</code></a> element of <a href=
"#term-SSML">SSML</a> (see <a href=
"http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/#S3.1.4">
Section 3.1.4</a> of [<a href="#ref-SSML">SSML</a>]).</p>
<p>The following is a simple example of an <a href=
"#term-SSML">SSML</a> document. It includes an Italian movie
title and the name of the director to be read in US English.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
xmlns="http://www.w3.org/2001/10/synthesis"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
xml:lang="en-US"&gt;
The title of the movie is: "La vita è bella" (Life is beautiful),
which is directed by Roberto Benigni.
&lt;/speak&gt;
</pre>
<p>To be pronounced correctly the Italian title and the
director's name might include the pronunciation inline in the
<a href="#term-SSML">SSML</a> document.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
xmlns="http://www.w3.org/2001/10/synthesis"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
xml:lang="en-US"&gt;
The title of the movie is:
&lt;phoneme alphabet="ipa" ph="ˈlɑ ˈviːɾə ˈʔeɪ ˈbɛlə"&gt;"La vita è bella"&lt;/phoneme&gt;
&lt;!-- The IPA pronunciation is:
"&amp;#x02C8;l&amp;#x0251; &amp;#x02C8;vi&amp;#x02D0;&amp;#x027E;&amp;#x0259;
&amp;#x02C8;&amp;#x0294;e&amp;#x026A; &amp;#x02C8;b&amp;#x025B;l&amp;#x0259;" --&gt;
(Life is beautiful),
which is directed by
&lt;phoneme alphabet="ipa" ph="ɹəˈbɛːɹɾoʊ bɛˈniːnji"&gt;Roberto Benigni.&lt;/phoneme&gt;
&lt;!-- The IPA pronunciation is:
"&amp;#x0279;&amp;#x0259;&amp;#x02C8;b&amp;#x025B;&amp;#x02D0;&amp;#x0279;&amp;#x027E;o&amp;#x028A;
b&amp;#x025B;&amp;#x02C8;ni&amp;#x02D0;nji" --&gt;
&lt;/speak&gt;
</pre>
<p>Using PLS, all the pronunciations can be factored out into an
external PLS document which is referenced by the <a href=
"#S4.4"><code>&lt;lexicon&gt;</code></a> element of <a href=
"#term-SSML">SSML</a> (see <a href=
"http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/#S3.1.4">
Section 3.1.4</a> of [<a href="#ref-SSML">SSML</a>]).</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
xmlns="http://www.w3.org/2001/10/synthesis"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
xml:lang="en-US"&gt;
&lt;lexicon uri="http://www.example.com/movie_lexicon.pls"/&gt;
The title of the movie is: "La vita è bella" (Life is beautiful),
which is directed by Roberto Benigni.
&lt;/speak&gt;
</pre>
<p>The referenced lexicon might look something like this:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;La vita &amp;#x00E8;&lt;!-- same as: è --&gt; bella&lt;/grapheme&gt;
&lt;phoneme&gt;ˈlɑ ˈviːɾə ˈʔeɪ ˈbɛlə&lt;/phoneme&gt;
&lt;!-- IPA string is:
"&amp;#x02C8;l&amp;#x0251; &amp;#x02C8;vi&amp;#x02D0;&amp;#x027E;&amp;#x0259;
&amp;#x02C8;&amp;#x0294;e&amp;#x026A; &amp;#x02C8;b&amp;#x025B;l&amp;#x0259;" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Roberto&lt;/grapheme&gt;
&lt;phoneme&gt;ɹəˈbɛːɹɾoʊ&lt;/phoneme&gt;
&lt;!-- IPA string is:
"&amp;#x0279;&amp;#x0259;&amp;#x02C8;b&amp;#x025B;&amp;#x02D0;&amp;#x0279;&amp;#x027E;o&amp;#x028A;" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Benigni&lt;/grapheme&gt;
&lt;phoneme&gt;bɛˈniːnji&lt;!-- IPA string is:
"b&amp;#x025B;&amp;#x02C8;ni&amp;#x02D0;nji" --&gt;&lt;/phoneme&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>The PLS engine will load the external PLS document and
transparently apply the pronunciations during the processing of
the <a href="#term-SSML">SSML</a> document. An application may
contain several distinct PLS documents to be used at different
points within the application. <a href=
"http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/#S3.1.4">
Section 3.1.4</a> of [<a href="#ref-SSML">SSML</a>] describes how
to use more than one lexicon document referenced in a <a href=
"#term-SSML">SSML</a> document.</p>
<h4 id="S1.1.0">Informative Note:</h4>
<p>Given that many platform/browser/text editor combinations do
not correctly cut and paste Unicode text, IPA symbols may be
entered as numeric character references (see Section 4.1 on
Character and Entity References of either XML 1.0 [<a href=
"#ref-xml-10">XML10</a>] or XML 1.1 [<a href=
"#ref-xml-11">XML11</a>]) in the pronunciation. However, the
UTF-8 representation of an IPA symbol should always be used in
preference to its numeric character reference. In order to
overcome potential problems with viewing the UTF-8 representation
of IPA symbols in this document, pronunciation examples are also
shown in a comment using numeric character references.</p>
<h2 id="S1.2">1.2. How ASR Uses PLS</h2>
<p>An <a href="#term-ASR">ASR</a> engine transforms an audio
signal into a recognized sequence of words or a semantic
representation of the meaning of the utterance (see <a href=
"#term-SISR">Semantic Interpretation for Speech Recognition</a>
[<a href="#ref-SISR">SISR</a>] for a standard definition of
Semantic Interpretation).</p>
<p>An <a href="#term-ASR">ASR</a> grammar is used to improve
<a href="#term-ASR">ASR</a> performance by describing the
possible words and phrases the <a href="#term-ASR">ASR</a> might
recognize. <a href="#term-SRGS">SRGS</a> is the standard
definition of <a href="#term-ASR">ASR</a> grammars (see [<a href=
"#ref-SRGS">SRGS</a>] for details).</p>
<p>PLS may be used by an <a href="#term-ASR">ASR</a> processor to
allow multiple pronunciations of words and phrases, and also to
do limited text normalization, such as <a href=
"#term-acronym-exp">acronym expansion</a> and abbreviations.</p>
<p>PLS entries are applied to the graphemes inside <a href=
"#term-SRGS">SRGS</a> grammar rules to convert them into the
phonemes to be recognized. See the example below and the example
in <a href="#S1.3">Section 1.3</a> for a PLS document used for
both <a href="#term-ASR">ASR</a> and <a href=
"#term-TTS">TTS</a>.</p>
<p>There might be other uses of PLS, for instance in a dictation
system or for unconstrained <a href="#term-ASR">ASR</a>, which
might be beyond the scope of this specification.</p>
<p>This is a very simple <a href="#term-SRGS">SRGS</a> grammar
that allows the recognition of sentences like "Boston
Massachusetts" or "Miami Florida".</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;grammar version="1.0"
xmlns="http://www.w3.org/2001/06/grammar"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/06/grammar
http://www.w3.org/TR/speech-grammar/grammar.xsd"
xml:lang="en-US" root="city_state" mode="voice"&gt;
&lt;rule id="city" scope="public"&gt;
&lt;one-of&gt; &lt;item&gt;Boston&lt;/item&gt;
&lt;item&gt;Miami&lt;/item&gt;
&lt;item&gt;Fargo&lt;/item&gt; &lt;/one-of&gt;
&lt;/rule&gt;
&lt;rule id="state" scope="public"&gt;
&lt;one-of&gt; &lt;item&gt;Florida&lt;/item&gt;
&lt;item&gt;North Dakota&lt;/item&gt;
&lt;item&gt;Massachusetts&lt;/item&gt; &lt;/one-of&gt;
&lt;/rule&gt;
&lt;rule id="city_state" scope="public"&gt;
&lt;ruleref uri="#city"/&gt; &lt;ruleref uri="#state"/&gt;
&lt;/rule&gt;
&lt;/grammar&gt;
</pre>
<p>If a <a href="#term-Pron-Lexicon">pronunciation lexicon</a> is
referenced by a <a href="#term-SRGS">SRGS</a> grammar it can
allow multiple pronunciations of the word in the grammar to
accommodate different speaking styles. Here is the same grammar
with a reference to an external PLS document.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;grammar version="1.0"
xmlns="http://www.w3.org/2001/06/grammar"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/06/grammar
http://www.w3.org/TR/speech-grammar/grammar.xsd"
xml:lang="en-US" root="city_state" mode="voice"&gt;
&lt;lexicon uri="http://www.example.com/city_lexicon.pls"/&gt;
&lt;rule id="city" scope="public"&gt;
&lt;one-of&gt; &lt;item&gt;Boston&lt;/item&gt;
&lt;item&gt;Miami&lt;/item&gt;
&lt;item&gt;Fargo&lt;/item&gt; &lt;/one-of&gt;
&lt;/rule&gt;
&lt;rule id="state" scope="public"&gt;
&lt;one-of&gt; &lt;item&gt;Florida&lt;/item&gt;
&lt;item&gt;North Dakota&lt;/item&gt;
&lt;item&gt;Massachusetts&lt;/item&gt; &lt;/one-of&gt;
&lt;/rule&gt;
&lt;rule id="city_state" scope="public"&gt;
&lt;ruleref uri="#city"/&gt; &lt;ruleref uri="#state"/&gt;
&lt;/rule&gt;
&lt;/grammar&gt;
</pre>
<p>Note also that an <a href="#term-SRGS">SRGS</a> grammar might
reference multiple PLS documents.</p>
<h2 id="S1.3">1.3. How VoiceXML Applications Use PLS</h2>
<p>A VoiceXML 2.0 application ([<a href="#ref-VXML">VXML</a>])
contains <a href="#term-SRGS">SRGS</a> grammars for <a href=
"#term-ASR">ASR</a> and <a href="#term-SSML">SSML</a> prompts for
<a href="#term-TTS">TTS</a>. The introduction of PLS into both
<a href="#term-SRGS">SRGS</a> and <a href="#term-SSML">SSML</a>
will directly impact VoiceXML applications.</p>
<p>The benefits described in <a href="#S1.1">Section 1.1</a> and
<a href="#S1.2">Section 1.2</a> are also available in VoiceXML
applications. The application may use several contextual PLS
documents at different points in the interaction, but may also
use the same PLS document both in <a href="#term-SRGS">SRGS</a>,
to improve <a href="#term-ASR">ASR</a>, and in <a href=
"#term-SSML">SSML</a>, to improve <a href="#term-TTS">TTS</a>.
Here is an example PLS document:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;judgment&lt;/grapheme&gt;
&lt;grapheme&gt;judgement&lt;/grapheme&gt;
&lt;phoneme&gt;ˈdʒʌdʒ.mənt&lt;/phoneme&gt;
&lt;!-- IPA string is:
"&amp;#x02C8;d&amp;#x0292;&amp;#x028C;d&amp;#x0292;&amp;#x002E;m&amp;#x0259;nt" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;fiancé&lt;/grapheme&gt;
&lt;grapheme&gt;fiance&lt;/grapheme&gt;
&lt;phoneme&gt;fiˈɒns.eɪ&lt;/phoneme&gt;
&lt;!-- IPA string is:
"fi&amp;#x02C8;&amp;#x0252;ns&amp;#x002E;e&amp;#x026A;" --&gt;
&lt;phoneme&gt;ˌfiː.ɑːnˈseɪ&lt;/phoneme&gt;
&lt;!-- IPA string is:
"&amp;#x02CC;fi&amp;#x02D0;&amp;#x002E;&amp;#x0251;&amp;#x02D0;n&amp;#x02C8;se&amp;#x026A;" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>which could be used to improve <a href="#term-TTS">TTS</a> as
shown in the following <a href="#term-SSML">SSML</a>
document:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
xmlns="http://www.w3.org/2001/10/synthesis"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
xml:lang="en-US"&gt;
&lt;lexicon uri="http://www.example.com/lexicon_defined_above.xml"/&gt;
&lt;p&gt; In the judgement of my fiancé, Las Vegas is the best place for a honeymoon.
I replied that I preferred Venice and didn't think the Venetian casino was an
acceptable compromise.&lt;\p&gt;
&lt;/speak&gt;
</pre>
<p>but also to improve <a href="#term-ASR">ASR</a> in the
following <a href="#term-SRGS">SRGS</a> grammar:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;grammar version="1.0"
xmlns="http://www.w3.org/2001/06/grammar"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/06/grammar
http://www.w3.org/TR/speech-grammar/grammar.xsd"
xml:lang="en-US" root="movies" mode="voice"&gt;
&lt;lexicon uri="http://www.example.com/lexicon_defined_above.xml"/&gt;
&lt;rule id="movies" scope="public"&gt;
&lt;one-of&gt;
&lt;item&gt;Terminator 2: Judgment Day&lt;/item&gt;
&lt;item&gt;My Big Fat Obnoxious Fiance&lt;/item&gt;
&lt;item&gt;Pluto's Judgement Day&lt;/item&gt;
&lt;/one-of&gt;
&lt;/rule&gt;
&lt;/grammar&gt;
</pre>
<h2 id="S1.4">1.4. What PLS does not Support</h2>
<p>The current specification is focused on the major features
described in the requirements document [<a href=
"#ref-REQS">REQS</a>]. The most complex features have been
postponed to a future revision of this specification. Some of the
complex features not included, for instance, are the introduction
of morphological, syntactic and semantic information associated
with pronunciations (such as word stems, inter-word semantic
links, pronunciation statistics, etc.). Many of these features
can be specified using RDF [<a href=
"#ref-rdf-xml">RDF-XMLSYNTAX</a>] that reference lexemes within
one or more <a href="#term-Pron-Lexicon">pronunciation
lexicons</a>.</p>
<h2 id="S1.5">1.5. Glossary of Terms</h2>
<dl>
<dt><b><a id="term-acronym" name=
"term-acronym">Acronym</a></b></dt>
<dd>A word formed from the initial letters of a name, such as
<b><i>PC</i></b> for <b>P</b>ersonal <b>C</b>omputer, or by
combining initial letters or parts of a series of words, such
as <b><i>radar</i></b> for <b>ra</b>dio <b>d</b>etection
<b>a</b>nd <b>r</b>anging, or variations, such as
<b><i>W3C</i></b> for <b>W</b>orld <b>W</b>ide <b>W</b>eb
<b>c</b>onsortium.</dd>
<dt><b><a id="term-acronym-exp" name="term-acronym-exp">Acronym
expansion</a></b></dt>
<dd>The action of replacing an <a href=
"#term-acronym">acronym</a> by the sequence of words it
represents. Acronym expansion is typically performed to help
<a href="#term-TTS">TTS</a> engines read acronyms and <a href=
"#term-ASR">ASR</a> to recognize them.</dd>
<dt><b><a id="term-ASR" name="term-ASR">ASR, Automatic Speech
Recognition, Speech Recognition</a></b></dt>
<dd>The process of using an automatic computation algorithm to
analyze spoken utterances to determine what words and phrases
or semantic information were present.</dd>
<dt><b><a id="term-grapheme" name=
"term-grapheme">Grapheme</a></b></dt>
<dd>One of the set of the smallest units of a written language,
such as letters, ideograms, or symbols, that distinguish one
word from another; a representation of a single orthographic
element.</dd>
<dt><b><a id="term-Homophone" name=
"term-Homophone">Homophone</a></b></dt>
<dd>One of a set of words that are pronounced the same way but
differ in meaning, origin, and sometimes spelling. E.g.
<i>night</i> and <i>knight</i> in English. Note that
<i>color</i> and <i>colour</i> are considered multiple <a href=
"#term-Orthography">orthographies</a> for the same word and not
homophones, because they are variations of spelling with the
same pronunciation and meaning.</dd>
<dt><b><a id="term-Homograph" name=
"term-Homograph">Homograph</a></b></dt>
<dd>'a word of the same written form as another but of
different meaning and usually origin, whether pronounced the
same way or not, as <em>bear</em> "to carry; support" and
<em>bear</em> "animal" or <em>lead</em> "to conduct" and
<em>lead</em> "metal."' [<a href="#ref-dict">DICT</a>] An
example from French is <i>fils</i> (son) and <i>fils</i>
(threads).</dd>
<dt><b><a id="term-IPA" name="term-IPA">International Phonetic
Alphabet (IPA)</a></b></dt>
<dd>The International Phonetic Alphabet [<a href=
"#ref-ipa">IPA</a>] is a <a href=
"#term-Phonetic-Alphabet">phonetic alphabet</a> used by
linguists to accurately and uniquely represent each of the wide
variety of sounds (phones or <a href=
"#term-Phoneme">phonemes</a>) the human vocal apparatus can
produce. It is intended as a notational standard for the
phonetic representation of all languages.</dd>
<dt><b><a id="term-Lexeme" name=
"term-Lexeme">Lexeme</a></b></dt>
<dd>An atomic unit in a language, like a word or a stem. In
this specification, "lexeme" designates a collection of
<a href="#term-grapheme">graphemic</a> and pronunciation
representations (e.g. <a href="#term-IPA">IPA</a>, <a href=
"#term-SAMPA">SAMPA</a>, Pinyin, etc.) of words or
phrases.</dd>
<dt><b><a id="term-Lexicon" name=
"term-Lexicon">Lexicon</a></b></dt>
<dd>In its most general sense, a lexicon is merely a list of
words or phrases, possibly containing information associated
with and related to the items in the list. This document uses
the term "lexicon" in only one specific way, as <a href=
"#term-Pron-Lexicon">"pronunciation lexicon"</a>, which means a
mapping between words (or short phrases), their written
representations, and their pronunciations suitable for use by
an <a href="#term-ASR">ASR</a> engine or a <a href=
"#term-TTS">TTS</a> engine. However, the word "lexicon" can
mean other things in other contexts.</dd>
<dt><b><a id="term-Orthography" name=
"term-Orthography">Orthography</a></b></dt>
<dd>A notation for writing and displaying words. Orthography
includes character sets, white space, case sensitivity,
diacritics within languages such as Arabic or Persian, and
accents within languages such as French.</dd>
<dt><b><a id="term-Phoneme" name=
"term-Phoneme">Phoneme</a></b></dt>
<dd>One of the set of the smallest units of speech that can
distinguish words: for example, the English language has over
40 <a href="#term-Phoneme">phonemes</a> (19 vowels and 24
consonants). In American English, /t/ and /p/ are <a href=
"#term-Phoneme">phonemes</a> that can distinguish the word
<i>tin</i> from <i>pin</i>.</dd>
<dt><b><a id="term-Phonetic-Alphabet" name=
"term-Phonetic-Alphabet">Phonetic alphabet</a></b></dt>
<dd>A set of symbols that represent the sounds in spoken
languages, such as English, Chinese, or German, see also
<a href="#term-IPA">International Phonetic Alphabet</a>.</dd>
<dt><b><a id="term-Pron-Lexicon" name=
"term-Pron-Lexicon">Pronunciation lexicon</a></b></dt>
<dd>The term pronunciation lexicon means a mapping between
words (or short phrases), their written representations, and
their pronunciations suitable for use by an <a href=
"#term-ASR">ASR</a> engine or a <a href="#term-TTS">TTS</a>
engine.</dd>
<dt><b><a id="term-SAMPA" name="term-SAMPA">SAMPA</a></b></dt>
<dd>The Speech Assessment Methods Phonetic Alphabet [<a href=
"#ref-sampa">SAMPA</a>]: a <a href=
"#term-Phonetic-Alphabet">phonetic alphabet</a> using only
ASCII characters, rather than the extended character set used
by the <a href="#term-IPA">International Phonetic
Alphabet</a>.</dd>
<dt><b><a id="term-SISR" name="term-SISR">Semantic
Interpretation for Speech Recognition</a> [<a href=
"#ref-SISR">SISR</a>]</b></dt>
<dd>A W3C specification defining a process to produce a
semantic result representing the meaning of a natural language
utterance.</dd>
<dt><b><a id="term-SRGS" name="term-SRGS">Speech Recognition
Grammar Specification</a> [<a href=
"#ref-SRGS">SRGS</a>]</b></dt>
<dd>A W3C specification defining a language to describe
grammars (words or phrases) that an <a href="#term-ASR">ASR</a>
engine can recognize.</dd>
<dt><b><a id="term-SSML" name="term-SSML">Speech Synthesis
Markup Language</a> [<a href="#ref-SSML">SSML</a>]</b></dt>
<dd>A W3C XML language for specifying the rendering of text by
a <a href="#term-TTS">TTS</a> engine.</dd>
<dt><b><a id="term-TTS" name="term-TTS">TTS, Text-To-Speech,
Speech Synthesis</a></b></dt>
<dd>Converting text into sounds using speech synthesis
techniques.</dd>
<dt><b><em><a id="term-uri" name="term-uri" shape="rect" href=
"http://www.w3.org/TR/webarch/#def-uri">URI: Uniform Resource
Identifier</a></em></b></dt>
<dd>A global identifier in the context of the World Wide Web
[<a href="#ref-web-arch" shape="rect">WEB-ARCH</a>]. A URI is
defined as any legal <code>anyURI</code> primitive as defined
in Section 3.2.17 of XML Schema Part 2: Datatypes [<a href=
"#ref-xmlschema-2" shape="rect">XML-SCHEMA-2</a>]. For
informational purposes only, [<a href="#ref-rfc3986" shape=
"rect">RFC3986</a>] and [<a href="#ref-rfc2732" shape=
"rect">RFC2732</a>] may be useful in understanding the
structure, format, and use of URIs. Note that IRIs (see
[<a href="#ref-rfc3987" shape="rect">RFC3987</a>]) are
permitted within the above definition of URI.</dd>
</dl>
<h2><a id="S2" name="S2"></a>2. Pronunciation Alphabets</h2>
<p>A phonemic/<a href="#term-Phonetic-Alphabet">phonetic
alphabet</a> is used to specify a pronunciation. An alphabet in
this context refers to a collection of symbols to represent the
sounds of one or more human languages. In the PLS specification
the pronunciation alphabet is specified by the
<code>alphabet</code> attribute (see <a href="#S4.1">Section
4.1</a> and <a href="#S4.6">Section 4.6</a> for details on the
use of this attribute). The only valid values for the
<code>alphabet</code> attribute are <code>"ipa"</code> (see the
next paragraph) and vendor-defined strings of the form
<code>"x-organization"</code> or
<code>"x-organization-alphabet"</code>. For example, the Japan
Electronics and Information Technology Industries Association
[<a href="#ref-jeita">JEITA</a>] might wish to encourage the use
of an alphabet such as <code>"x-jeita"</code> or
<code>"x-jeita-2000"</code> for their <a href=
"#term-Phoneme">phoneme</a> alphabet [<a href=
"#ref-jeidaalphabet">JEIDAALPHABET</a>]. Another example might be
<code>"x-sampa"</code> [<a href="#ref-x-sampa">X-SAMPA</a>], an
extension of the <a href="#term-SAMPA">SAMPA</a> phonetic
alphabet [<a href="#ref-sampa">SAMPA</a>] to cover the entire
range of characters in the <a href="#term-IPA">International
Phonetic Alphabet</a> [<a href="#ref-ipa">IPA</a>].</p>
<p>A compliant PLS processor <em title="MUST in RFC 2119 context"
class="RFC2119">MUST</em> support <code>"ipa"</code> as the value
of the <code>alphabet</code> attribute. This means that the PLS
processor <em title="MUST in RFC 2119 context" class=
"RFC2119">MUST</em> support the Unicode representations of the
phonetic characters developed by the International Phonetic
Association [<a href="#ref-ipa">IPA</a>]. In addition to an
exhaustive set of vowel and consonant symbols, this character set
supports a syllable delimiter, numerous diacritics, stress
symbols, lexical tone symbols, intonational markers and more. For
this alphabet, legal phonetic/phonemic values are strings of the
values specified in Appendix 2 of [<a href=
"#ref-ipahndbk">IPAHNDBK</a>]; note that an IPA transcription may
contain white space characters to assist readability, which have
no implications for the pronunciation. Informative tables of the
IPA-to-Unicode mappings can be found at [<a href=
"#ref-ipaunicode1">IPAUNICODE1</a>] and [<a href=
"#ref-ipaunicode2">IPAUNICODE2</a>]. Note that not all of the IPA
characters are available in Unicode. For processors supporting
this alphabet,</p>
<ul>
<li>The processor <em title="MUST in RFC 2119 context" class=
"RFC2119">MUST</em> syntactically accept all legal values.</li>
<li>The processor <em title="SHOULD in RFC 2119 context" class=
"RFC2119">SHOULD</em> handle all Unicode IPA codes that can
reasonably be considered to belong to the current
language.</li>
</ul>
<h4 id="S2.0">Informative Note:</h4>
<p>Note that there are peculiarities in the IPA alphabet which
might have implications for implementers, for instance
equivalent, withdrawn and superseded IPA symbols; see Appendix 2
of [<a href="#ref-ipahndbk">IPAHNDBK</a>] for further
details.</p>
<h4 id="S2.1">Informative Note:</h4>
<p>When IPA symbols are used to represent the phonemes of a
language, there can be an ambiguity concerning which allophonic
symbol to select to represent a phoneme. Note that this may
result in inconsistencies between lexicons which were composed
for the identical language.</p>
<h4 id="S2.2">Informative Note:</h4>
<p>Currently there is no ready way for a blind or partially
sighted person to read or interact with a lexicon containing IPA
symbols. It is hoped that implementers will provide tools which
will enable such an interaction.</p>
<div>
<h2 id="S3">3. PLS Documents</h2>
<h3 id="S3.1">3.1 Document Form</h3>
<p>A legal Pronunciation Lexicon Specification document
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
have a legal XML Prolog from Section 2.8 of either XML 1.0
[<a href="#ref-xml-10">XML10</a>] or XML 1.1 [<a href=
"#ref-xml-11">XML11</a>]. The XML prolog is followed by the
root <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element.
See <a href="#S4.1">Section 4.1</a> for details on this
element.</p>
<p>The <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
designate the PLS namespace. This can be achieved by declaring
an <code class="att">xmlns</code> attribute or an attribute
with an "xmlns" prefix. See Section 2 of Namespaces in XML
(Namespaces in XML 1.0 [<a href="#ref-xmlns-10">XML-NS10</a>]
or Namespaces in XML 1.1 [<a href=
"#ref-xmlns-11">XML-NS11</a>]) for details. Note that when the
<code class="att">xmlns</code> attribute is used alone, it sets
the default namespace for the element on which it appears and
for any child elements. The namespace for PLS is defined to be
<code>"http://www.w3.org/2005/01/pronunciation-lexicon"</code>.</p>
<p>It is <em title="RECOMMENDED in RFC 2119 context" class=
"RFC2119">RECOMMENDED</em> that the <a href=
"#S4.1"><code>&lt;lexicon&gt;</code></a> element also indicate
the location of the PLS schema (see <a href="#AppA">Appendix
A</a>) via the <code class="att"><a href=
"http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/#xsi_schemaLocation">
xsi:schemaLocation</a></code> attribute from <a href=
"http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/#Instance_Document_Constructions">
Section 2.6.3</a> of XML Schema Part 1: Structures Second
Edition [<a href="#ref-xmlschema-1">XML-SCHEMA-1</a>].</p>
<p>The following is an example of a legal PLS header:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
</pre>
</div>
<h3 id="S3.2">3.2 Conformance</h3>
<p>This section enumerates the conformance rules of this
specification.</p>
<p>All sections in this specification are normative, unless
otherwise indicated. The informative parts of this specification
are identified by "Informative" labels within sections.</p>
<p>Individual conformance requirements or testable statements are
identifiable in the PLS specification through imperative voice
statements. The key words "<em title="MUST in RFC 2119 context"
class="RFC2119">MUST</em>", "<em title=
"MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>",
"<em title="REQUIRED in RFC 2119 context" class=
"RFC2119">REQUIRED</em>", "<em title="SHALL in RFC 2119 context"
class="RFC2119">SHALL</em>", "<em title=
"SHALL NOT in RFC 2119 context" class="RFC2119">SHALL NOT</em>",
"<em title="SHOULD in RFC 2119 context" class=
"RFC2119">SHOULD</em>", "<em title=
"SHOULD NOT in RFC 2119 context" class="RFC2119">SHOULD
NOT</em>", "<em title="RECOMMENDED in RFC 2119 context" class=
"RFC2119">RECOMMENDED</em>", "<em title="MAY in RFC 2119 context"
class="RFC2119">MAY</em>", and "<em title=
"OPTIONAL in RFC 2119 context" class="RFC2119">OPTIONAL</em>" in
this document are to be interpreted as described in [<a href=
"#ref-rfc2119">RFC2119</a>]. However, for readability, these
words do not appear in all uppercase letters in this
specification.</p>
<div>
<h3 id="S3.2.1">3.2.1 Conforming Pronunciation Lexicon
Specification Documents</h3>
<p>A document is a <em>Conforming Pronunciation Lexicon
Specification Document</em> if it meets both the following
conditions:</p>
<ul>
<li>It is well-formed according to the version of XML used
(either XML 1.0 [<a href="#ref-xml-10">XML10</a>] or XML 1.1
[<a href="#ref-xml-11">XML11</a>]) and conforms to the
corresponding Namespaces in XML specification (Namespaces in
XML 1.0 [<a href="#ref-xmlns-10">XML-NS10</a>] or Namespaces
in XML 1.1 [<a href="#ref-xmlns-11">XML-NS11</a>]).</li>
<li>It adheres to the specification described in this
document (<a href="#S1" shape="rect">Pronunciation Lexicon
Specification</a>) including the constraints expressed in the
Schema (see <a href="#AppA" shape="rect">Appendix A</a>) and
having an XML Prolog and <a href=
"#S4.1"><code>&lt;lexicon&gt;</code></a> root element as
specified in <a href="#S3.1">Section 3.1</a>.</li>
</ul>
<p>This specification and these conformance criteria provide no
designated size limits on any aspect of PLS documents. There
are no maximum values on the number of elements, the amount of
character data, or the number of characters in attribute
values.</p>
<h3 id="S3.2.2">3.2.2 Using PLS with other Namespaces</h3>
<p>The PLS namespace <em title="MAY in RFC 2119 context" class=
"RFC2119">MAY</em> be used with other XML namespaces as per the
Namespaces in XML Recommendations (Namespaces in XML 1.0
[<a href="#ref-xmlns-10">XML-NS10</a>] or Namespaces in XML 1.1
[<a href="#ref-xmlns-11">XML-NS11</a>]). Future work by W3C is
expected to address ways to specify conformance for documents
involving multiple namespaces.</p>
<h3 id="S3.2.3">3.2.3 Conforming Pronunciation Lexicon
Specification Processors</h3>
<p>A <em>Conforming Pronunciation Lexicon Specification
Processor</em> <em title="MUST in RFC 2119 context" class=
"RFC2119">MUST</em> be able to parse and process <a href=
"#S3.2.1">Conforming Pronunciation Lexicon Specification
documents</a>.</p>
<p>In a <em>Conforming Pronunciation Lexicon Specification
Processor</em>, the XML parser <em title=
"MUST in RFC 2119 context" class="RFC2119">MUST</em> be able to
parse and process all XML constructs defined by either XML 1.0
[<a href="#ref-xml-10">XML10</a>] or XML 1.1 [<a href=
"#ref-xml-11">XML11</a>] and conforms to the corresponding
Namespaces in XML specification (Namespaces in XML 1.0
[<a href="#ref-xmlns-10">XML-NS10</a>] or Namespaces in XML 1.1
[<a href="#ref-xmlns-11">XML-NS11</a>]).</p>
<p>A Conforming Pronunciation Lexicon Specification Processor
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
conform to the XML 1.0 or XML 1.1 requirements for conformant
non validating processors.</p>
<p>A Conforming Pronunciation Lexicon Specification Processor
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
correctly understand and apply the semantics of each markup
element as described by this document.</p>
<p>A Conforming Pronunciation Lexicon Specification Processor
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
meet the following requirements for handling of natural (human)
languages:</p>
<ul>
<li>A Conforming Pronunciation Lexicon Specification
Processor is <em title="REQUIRED in RFC 2119 context" class=
"RFC2119">REQUIRED</em> to parse all legal natural (human)
language declarations successfully.</li>
<li>For any Conforming Pronunciation Lexicon Specification
Document whose <code>xml:lang</code> attribute (on the
<a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element) has
a value representing a natural (human) language that the
Pronunciation Lexicon Specification Processor claims to
support, the Processor is <em title=
"REQUIRED in RFC 2119 context" class="RFC2119">REQUIRED</em>
to successfully parse and treat all text encountered as if in
that language in order to be a Conforming Processor.</li>
<li>A Conforming Pronunciation Lexicon Specification
Processor <em title="SHOULD in RFC 2119 context" class=
"RFC2119">SHOULD</em> inform the hosting environment when it
parses a Conforming Pronunciation Lexicon Specification
Document whose <code>xml:lang</code> attribute (on the
<a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element) has
a value representing a natural (human) language that the
Processor does not support.</li>
</ul>
<p>When a Conforming Pronunciation Lexicon Specification
Processor encounters elements or attributes that are not
declared in this specification and such elements or attributes
occur where it is not forbidden in this specification, the
processor <em title="MAY in RFC 2119 context" class=
"RFC2119">MAY</em> choose to:</p>
<ul>
<li>ignore the non-standard elements and/or attributes</li>
<li>or, process the non-standard elements and/or
attributes</li>
<li>or, reject the document containing those elements and/or
attributes</li>
</ul>
<p>Except where stated in this document, there is no
conformance requirement with respect to performance of
rendering pronunciations as acoustic structures (models,
waveforms, etc.) for <a href="#term-ASR">ASR</a> and <a href=
"#term-TTS">TTS</a>.</p>
</div>
<h2 id="S4">4. Pronunciation Lexicon Markup Language
Definition</h2>
<p>The <a href="#term-Pron-Lexicon">Pronunciation Lexicon</a>
markup language consists of the following elements and
attributes:</p>
<table border="1" summary=
"Elements and Attributes for Pronunciation Lexicon">
<thead>
<tr>
<th width="25%" scope="col">Elements</th>
<th width="25%" scope="col">Attributes</th>
<th width="50%" scope="col">Description</th>
</tr>
</thead>
<tbody>
<tr>
<td><a href="#S4.1"><code>&lt;lexicon&gt;</code></a></td>
<td><code>version</code><br />
<code>xml:base</code><br />
<code>xmlns</code><br />
<code>xml:lang</code><br />
<code>alphabet</code></td>
<td>root element for PLS</td>
</tr>
<tr>
<td><a href="#S4.2"><code>&lt;meta&gt;</code></a></td>
<td><code>name</code><br />
<code>http-equiv</code><br />
<code>content</code></td>
<td>element containing meta data</td>
</tr>
<tr>
<td><a href="#S4.3"><code>&lt;metadata&gt;</code></a></td>
<td>&nbsp;</td>
<td>element containing meta data</td>
</tr>
<tr>
<td><a href="#S4.4"><code>&lt;lexeme&gt;</code></a></td>
<td><code>xml:id</code><br />
<code>role</code></td>
<td>the container element for a single lexical entry</td>
</tr>
<tr>
<td><a href="#S4.5"><code>&lt;grapheme&gt;</code></a></td>
<td>&nbsp;</td>
<td>contains <a href="#term-Orthography">orthographic
information</a> for a <a href=
"#term-Lexeme">lexeme</a></td>
</tr>
<tr>
<td><a href="#S4.6"><code>&lt;phoneme&gt;</code></a></td>
<td><code>prefer</code><br />
<code>alphabet</code></td>
<td>contains pronunciation information for a <a href=
"#term-Lexeme">lexeme</a></td>
</tr>
<tr>
<td><a href="#S4.7"><code>&lt;alias&gt;</code></a></td>
<td><code>prefer</code></td>
<td>contains <a href="#term-acronym-exp">acronym
expansions</a> and orthographic substitutions</td>
</tr>
<tr>
<td><a href="#S4.8"><code>&lt;example&gt;</code></a></td>
<td>&nbsp;</td>
<td>contains an example of the usage for a <a href=
"#term-Lexeme">lexeme</a></td>
</tr>
</tbody>
</table>
<h3 id="S4.1">4.1 <code>&lt;lexicon&gt;</code> Element</h3>
<p>The root element of the Pronunciation Lexicon markup language
is the <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element.
This element is the container for all other elements of the PLS
language. A <a href="#S4.1"><code>&lt;lexicon&gt;</code></a>
element <em title="MUST in RFC 2119 context" class=
"RFC2119">MUST</em> contain zero or more <a href=
"#S4.2"><code>&lt;meta&gt;</code></a> elements, followed by an
<em title="OPTIONAL in RFC 2119 context" class=
"RFC2119">OPTIONAL</em> <a href=
"#S4.3"><code>&lt;metadata&gt;</code></a> element, followed by
zero or more <a href="#S4.4"><code>&lt;lexeme&gt;</code></a>
elements. Note that a PLS document without any <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> elements may be useful as
a placeholder for future lexical entries.</p>
<p>The <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
specify an <code>alphabet</code> attribute which indicates the
default pronunciation alphabet to be used within the PLS
document. The values of the <code>alphabet</code> attribute are
described in <a href="#S2">Section 2</a>. The default
pronunciation alphabet <em title="MAY in RFC 2119 context" class=
"RFC2119">MAY</em> be overridden for a given lexeme using the
<a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element.</p>
<p>The <em title="REQUIRED in RFC 2119 context" class=
"RFC2119">REQUIRED</em> <code>version</code> attribute indicates
the version of the specification to be used for the document and
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
have the value <code>"1.0"</code>.</p>
<p>The <em title="REQUIRED in RFC 2119 context" class=
"RFC2119">REQUIRED</em> <code>xml:lang</code> attribute allows
identification of the language for which the <a href=
"#term-Pron-Lexicon">pronunciation lexicon</a> is relevant. IETF
Best Current Practice 47 [<a href="#ref-bcp47" shape=
"rect">BCP47</a>] is the normative reference on the values of the
<code>xml:lang</code> attribute.</p>
<p>Note that <code>xml:lang</code> specifies a single unique
language for the entire PLS document. This does not limit the
ability to create multilingual <a href="#term-SRGS">SRGS</a>
[<a href="#ref-SRGS">SRGS</a>] and <a href="#term-SSML">SSML</a>
[<a href="#ref-SSML">SSML</a>] documents. These documents may
reference multiple <a href="#term-Pron-Lexicon">pronunciation
lexicons</a>, possibly written for different languages.</p>
<p>The namespace URI for PLS is
<code>"http://www.w3.org/2005/01/pronunciation-lexicon"</code>.
All PLS markup <em title="MUST in RFC 2119 context" class=
"RFC2119">MUST</em> be associated with the PLS namespace, using a
Namespace Declaration as described in either Namespaces in XML
1.0 [<a href="#ref-xmlns-10">XML-NS10</a>] or Namespaces in XML
1.1 [<a href="#ref-xmlns-11">XML-NS11</a>]. This can for instance
be achieved by declaring an <code>xmlns</code> attribute on the
<a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element, as the
examples in this specification show.</p>
<p>PLS documents <em title="MAY in RFC 2119 context" class=
"RFC2119">MAY</em> include the <code>xml:base</code> attribute as
defined in [<a href="#ref-xml-base" shape="rect">XML-BASE</a>].
Note that as in the HTML 4.01 specification [<a href="#ref-html"
shape="rect">HTML</a>], this is a URI which all the relative
references within the document take as their base.</p>
<h4 id="S4.1.0.0">Informative Note:</h4>
<p>Note that in this version of the specification, only the
contents of metadata can potentially use relative URIs.</p>
<h4 id="S4.1.1">Example:</h4>
<p>A simple PLS document for the word "tomato" and its
pronunciation.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;tomato&lt;/grapheme&gt;
&lt;phoneme&gt;təmei̥ɾou̥&lt;/phoneme&gt;
&lt;!-- IPA string is: "t&amp;#x0259;mei&amp;#x325;&amp;#x027E;ou&amp;#x325;" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S4.2">4.2 <code>&lt;meta&gt;</code> Element</h3>
<p>The <a href="#S4.3"><code>&lt;metadata&gt;</code></a> and
<a href="#S4.2"><code>&lt;meta&gt;</code></a> elements are
containers in which information about the document can be placed.
The <a href="#S4.3"><code>&lt;metadata&gt;</code></a> element
provides more general and powerful treatment of metadata
information than <a href="#S4.2"><code>&lt;meta&gt;</code></a> by
using a metadata schema.</p>
<p>A <a href="#S4.2"><code>&lt;meta&gt;</code></a> element
associates a string to a declared meta property or declares
<code>http-equiv</code> content. Either a <code>name</code> or
<code>http-equiv</code> attribute is <em title=
"REQUIRED in RFC 2119 context" class="RFC2119">REQUIRED</em>. It
is an error to provide both <code>name</code> and
<code>http-equiv</code> attributes. A <code>content</code>
attribute is also <em title="REQUIRED in RFC 2119 context" class=
"RFC2119">REQUIRED</em>. The only <a href=
"#S4.2"><code>&lt;meta&gt;</code></a> property defined by this
specification is <code>"seeAlso"</code>. It is used to specify a
resource that might provide additional metadata information about
the content. This property is modeled on the
<code>"seeAlso"</code> property from <a href=
"http://www.w3.org/TR/2004/REC-rdf-schema-20040210/#ch_seealso">Section
5.4.1</a> of "RDF Vocabulary Description Language 1.0: RDF
Schema" [<a href="#ref-rdf-schema">RDF-SCHEMA</a>]. The
<code>http-equiv</code> attribute has a special significance when
documents are retrieved via HTTP. Although the preferred method
of providing HTTP header information is to use HTTP header
fields, the <code>http-equiv</code> content <em title=
"MAY in RFC 2119 context" class="RFC2119">MAY</em> be used in
situations where the PLS document author is unable to configure
HTTP header fields associated with their document on the origin
server, for example, cache control information. Note that HTTP
servers and caches are not required to inspect the contents of
<a href="#S4.2"><code>&lt;meta&gt;</code></a> in PLS documents
and thereby override the header values they would send
otherwise.</p>
<p>The <a href="#S4.2"><code>&lt;meta&gt;</code></a> element is
an empty element.</p>
<h4 id="S4.2.0">Informative Note:</h4>
<p>This section is modeled after the <code>&lt;meta&gt;</code>
description in the HTML 4.01 Specification [<a href=
"#ref-html">HTML</a>]. Despite the fact that the name/content
model is now being replaced by better ways to include metadata,
see for instance <a href=
"http://www.w3.org/TR/2004/WD-xhtml2-20040722/mod-meta.html#sec_20.6.">
Section 20.6</a> of XHTML 2.0 [<a href="#ref-XHTML2">XHTML2</a>],
and the fact that the <code>http-equiv</code> directive is no
longer recommended in <a href=
"http://www.w3.org/TR/xhtml-media-types/xhtml-media-types.html#application-xml">
Section 3.3</a> of XHTML Media Types [<a href=
"#ref-XHTML-MTYPES">XHTML-MTYPES</a>], the Working Group has
decided to retain this for compatibility with the other
specifications of the first version of the Speech Interface
Framework (VoiceXML, SSML, SRGS, CCXML). Future versions of the
framework will align with more modern metadata schemes.</p>
<h4 id="S4.2.1">Example:</h4>
<p>This is an example of how <a href=
"#S4.2"><code>&lt;meta&gt;</code></a> elements can be included in
a PLS document to specify a resource that provides additional
metadata information.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;meta http-equiv="Cache-Control" content="no-cache"/&gt;
&lt;meta name="seeAlso" content="http://example.com/my-pls-metadata.xml"/&gt;
&lt;!-- If lexemes are to be added to this lexicon, they start below --&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S4.3">4.3 <code>&lt;metadata&gt;</code> Element</h3>
<p>The <a href="#S4.3"><code>&lt;metadata&gt;</code></a> element
is a container in which information about the document can be
placed using metadata markup. The behavior of software processing
the content of a <a href=
"#S4.3"><code>&lt;metadata&gt;</code></a> element is not
described in this specification. Therefore, software implementing
this specification is free to ignore that content.</p>
<p>Although any metadata markup can be used within <a href=
"#S4.3"><code>&lt;metadata&gt;</code></a>, it is <em title=
"RECOMMENDED in RFC 2119 context" class=
"RFC2119">RECOMMENDED</em> that the RDF/XML Syntax [<a href=
"#ref-rdf-xml">RDF-XMLSYNTAX</a>] be used, in conjunction with
the general metadata properties defined by the Dublin Core
Metadata Initiative [<a href="#ref-dc">DC</a>] (e.g., Title,
Creator, Subject, Description, Rights, etc.)</p>
<h4 id="S4.3.1">Example:</h4>
<p>This is an example of how metadata can be included in a PLS
document using the "Dublin Core Metadata Element Set, Version
1.1" [<a href="#ref-dc-es">DC-ES</a>] describing general document
information such as title, description, date, and so on:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;metadata&gt;
&lt;rdf:RDF
xmlns:rdf = "http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:dc = "http://purl.org/dc/elements/1.1/"&gt;
&lt;!-- Metadata about the PLS document --&gt;
&lt;rdf:Description rdf:about=""
dc:title="Pronunciation lexicon for W3C terms"
dc:description="Common pronunciations for many W3C acronyms and abbreviations, i.e. I18N or WAI"
dc:publisher="W3C"
dc:date="2005-11-29"
dc:rights="Copyright 2002 W3C"
dc:format="application/pls+xml"&gt;
&lt;dc:creator&gt;The W3C Voice Browser Working Group&lt;/dc:creator&gt;
&lt;/rdf:Description&gt;
&lt;/rdf:RDF&gt;
&lt;/metadata&gt;
&lt;!-- If lexemes are to be added to this lexicon, they start below --&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S4.4">4.4 <code>&lt;lexeme&gt;</code> Element</h3>
<p>The <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element is
a container for a lexical entry which <em title=
"MAY in RFC 2119 context" class="RFC2119">MAY</em> include
multiple <a href="#term-Orthography">orthographies</a> and
multiple pronunciation information.</p>
<p>The <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
contains one or more <a href=
"#S4.5"><code>&lt;grapheme&gt;</code></a> elements, one or more
pronunciations (either by <a href=
"#S4.5"><code>&lt;phoneme&gt;</code></a> or <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> elements or a combination
of both), and zero or more <a href=
"#S4.8"><code>&lt;example&gt;</code></a> elements. The children
of the <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
<em title="MAY in RFC 2119 context" class="RFC2119">MAY</em>
appear in any order, but note that the order will have an impact
on the treatment of multiple pronunciations (see <a href=
"#S4.9">Section 4.9</a>).</p>
<p>The <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
has an <em title="OPTIONAL in RFC 2119 context" class=
"RFC2119">OPTIONAL</em> <code>xml:id</code> [<a href=
"#ref-xml-id">XML-ID</a>] attribute, allowing the element to be
referenced from other documents (through fragment identifiers or
XPointer [<a href="#ref-xpointer">XPOINTER</a>], for instance).
For example, developers may use external RDF statements [<a href=
"#ref-rdf-conc">RDF-CONC</a>] to associate metadata (such as part
of speech or word relationships) with a lexeme.</p>
<p>The <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
has an <em title="OPTIONAL in RFC 2119 context" class=
"RFC2119">OPTIONAL</em> <code>role</code> attribute which takes
as its value one or more white space separated QNames as defined
in Section 4 of Namespaces in XML (1.0 [<a href=
"#ref-xmlns-10">XML-NS10</a>] or 1.1 [<a href=
"#ref-xmlns-11">XML-NS11</a>], depending on the version of XML
being used).</p>
<p>The <code>role</code> attribute describes additional
information to help the selection of the most relevant
pronunciation for a given <a href=
"#term-Orthography">orthography</a>. The main use is to
differentiate words that have the same spelling but are
pronounced in different ways (cf. <a href=
"#term-Homograph">homographs</a> and see also <a href=
"#S5.5">Section 5.5</a>). A QName in the attribute content of the
<code>role</code> attribute is expanded into an expanded-name
using the namespace declarations in scope for the containing
<a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element. Thus,
each QName provides a reference to a specific item in the
designated namespace. In the second example below, the QName
<code>"claws:VVI"</code> within the role attribute expands to the
<code>"VVI"</code> item in the
<code>"http://www.example.com/claws7tags"</code> namespace. This
mechanism allows for referencing defined taxonomies of word
classes, with the expectation that they are documented at the
specified namespace URI.</p>
<h4 id="S4.4.1">Example:</h4>
<p>A <a href="#term-Pron-Lexicon">pronunciation lexicon</a> for
the Italian language with two <a href="#term-Lexeme">lexemes</a>.
One of them is for the loan word "file" which is often used in
technical discussions to have the same meaning and pronunciation
as in English. This is distinct from the homograph noun "file"
which is the plural form of "fila" meaning "queue". Note that
this user-specified pronunciation for "file" takes precedence
over any system-defined pronunciation.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="it"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;file&lt;/grapheme&gt;
&lt;phoneme&gt;faɪl&lt;/phoneme&gt;
&lt;!-- This is the pronunciation
of the loan word "file" in Italian.
IPA string is: "fa&amp;#x026A;l" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;EU&lt;/grapheme&gt;
&lt;alias&gt;Unione Europea
&lt;!-- This is a substitution of the European
Union acronym in Italian language. --&gt;&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>The following is an example of a pronunciation lexicon for the
word "read":</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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"
xmlns:claws="http://www.example.com/claws7tags" alphabet="ipa"
xml:lang="en"&gt;
&lt;lexeme role="claws:VVI claws:VV0 claws:NN1"&gt;
&lt;!-- verb infinitive, verb present tense, singular noun --&gt;
&lt;grapheme&gt;read&lt;/grapheme&gt;
&lt;phoneme&gt;ri&amp;#x02D0;d&lt;!-- same as riːd --&gt;&lt;/phoneme&gt;
&lt;/lexeme&gt;
&lt;lexeme role="claws:VVN claws:VVD"&gt;
&lt;!-- verb past participle, verb past tense --&gt;
&lt;grapheme&gt;read&lt;/grapheme&gt;
&lt;phoneme&gt;red&lt;/phoneme&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>Note that the <code>role</code> attribute is based on
qualified values (in this example from the <a href=
"http://www.comp.lancs.ac.uk/ucrel/claws7tags.html">UCREL CLAWS7
tagset</a> of part-of-speech) to distinguish the verb infinitive,
present tense and singular noun from the verb past tense and past
participle pronunciation of the word "read".</p>
<p>The following is an example document which references the
above lexicon and includes an extension element to show how the
<code>role</code> attribute may be used to select the relevant
pronunciation of the word "read" in the dialog.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
xmlns="http://www.w3.org/2001/10/synthesis"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
xmlns:myssml="http://www.example.com/ssml_extensions"
xmlns:claws="http://www.example.com/claws7tags"
xml:lang="en"&gt;
&lt;lexicon uri="http://www.example.com/lexicon.pls"
type="application/pls+xml"/&gt;
&lt;voice gender="female" age="3"&gt;
Can you &lt;myssml:token role="claws:VVI"&gt;read&lt;/myssml:token&gt; this book
to me?
&lt;/voice&gt;
&lt;voice gender="male" age="43"&gt;
I've already &lt;myssml:token role="claws:VVN"&gt;read&lt;/myssml:token&gt; it
three times!
&lt;/voice&gt;
&lt;/speak&gt;
</pre>
<p>Here is another example in Chinese that uses SSML 1.1
[<a href="#ref-SSML-11">SSML-11</a>].</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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"
xmlns:claws="http://www.example.com/claws7tags"
alphabet="x-myorganization-pinyin"
xml:lang="zh-CN"&gt;
&lt;lexeme role="claws:VV0"&gt;
&lt;!-- base form of lexical verb --&gt;
&lt;grapheme&gt;&lt;/grapheme&gt;
&lt;phoneme&gt;chu3&lt;/phoneme&gt;
&lt;!-- pinyin string is: "chǔ" in 处罚 处置 --&gt;
&lt;/lexeme&gt;
&lt;lexeme role="claws:NN"&gt;
&lt;!-- common noun, neutral for number --&gt;
&lt;grapheme&gt;&lt;/grapheme&gt;
&lt;phoneme&gt;chu4&lt;/phoneme&gt;
&lt;!-- pinyin string is: "chù" in 处所 妙处 --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>This is a sample document which references the above lexicon
and shows how the <code>role</code> attribute may be used to
select the relevant pronunciation of the Chinese word "处" in the
dialog.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.1"
xmlns="http://www.w3.org/2001/10/synthesis"
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
xmlns:claws="http://www.example.com/claws7tags"
xml:lang="zh-CN"&gt;
&lt;lexicon uri="http://www.example.com/lexicon.pls"
type="application/pls+xml"
xml:id="mylex"/&gt;
&lt;lookup ref="mylex"&gt;
他这个人很不好相&lt;w role="claws:VV0"&gt;&lt;/w&gt;
&lt;w role="claws:NN"&gt;&lt;/w&gt;不准照相。
&lt;/lookup&gt;
&lt;/speak&gt;
</pre>
<h4 id="S4.4.2">Informative Note:</h4>
<p>The SRGS 1.0 [<a href="#ref-SRGS">SRGS</a>] and SSML 1.0
[<a href="#ref-SSML">SSML</a>] specifications do not currently
support a selection mechanism based on the <code>role</code>
attribute. Future versions of these specifications are expected
to allow the selection of relevant pronunciations on the basis of
the <code>role</code> attribute.</p>
<h3 id="S4.5">4.5 <code>&lt;grapheme&gt;</code> Element</h3>
<p>A <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> contains at
least one <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
element. The <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
element contains text describing the <a href=
"#term-Orthography">orthography</a> of the <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a>.</p>
<p>The <a href="#S4.5"><code>&lt;grapheme&gt;</code></a> element
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
contain 'character' child information items. The <a href=
"#S4.5"><code>&lt;grapheme&gt;</code></a> element <em title=
"MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>
contain 'element' child information items from any namespace,
i.e. PLS or foreign namespace.</p>
<p>In more complex situations there may be alternative textual
representations for the same word or phrase; this can arise due
to a number of reasons, for example:</p>
<ul class="noindent">
<li>Regional spelling variations e.g. "colour" and
"color";</li>
<li>Free spelling variations e.g. "judgment" and
"judgement"</li>
<li>Alternate writing systems, e.g. Japanese uses a mixture of
Han ideographs (Kanji), and phonemic spelling systems (Katakana
or Hiragana) for representing the <a href=
"#term-Orthography">orthography</a> of a word or phrase, and
such mixture sometimes has several variations as in kana
suffixes following kanji stems (Okurigana) for example "okonau"
(行なう vs. 行う);</li>
<li>Reformed spellings e.g. in German some of the words which
used to have "ß" before are now to be written with "ss".</li>
</ul>
<p>In order to remove the need for duplication of pronunciation
information to cope with the above variations, the <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> element <em title=
"MAY in RFC 2119 context" class="RFC2119">MAY</em> contain more
than one <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
element to define the base <a href=
"#term-Orthography">orthography</a> and any variants. Note that
all the pronunciations given within the <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> apply to each and every
<a href="#S4.5"><code>&lt;grapheme&gt;</code></a> within the
<a href="#S4.4"><code>&lt;lexeme&gt;</code></a>.</p>
<h4 id="S4.5.1">Examples:</h4>
<p>An example of a single <a href="#term-grapheme">grapheme</a>
and a single pronunciation.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Sepulveda&lt;/grapheme&gt;
&lt;phoneme&gt;səˈpʌlvɪdə&lt;/phoneme&gt;
&lt;!-- IPA string is: "s&amp;#x0259;&amp;#x02C8;p&amp;#x028C;lv&amp;#x026A;d&amp;#x0259;" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>Another example with more than one written form for a lexical
entry, where the first <a href=
"#term-Orthography">orthography</a> uses Latin characters for
"Romaji" <a href="#term-Orthography">orthography</a>, the second
one uses "Kanji" <a href="#term-Orthography">orthography</a> and
the third one uses the "Hiragana" <a href=
"#term-Orthography">orthography</a>:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="jp"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;nihongo&lt;!-- "Romaji" --&gt;&lt;/grapheme&gt;
&lt;grapheme&gt;日本語&lt;!-- "Kanji" --&gt;&lt;/grapheme&gt;
&lt;grapheme&gt;にほんご&lt;!-- "Hiragana" --&gt;&lt;/grapheme&gt;
&lt;phoneme&gt;ɲihoŋo
&lt;!-- IPA string is: "&amp;#x0272;iho&amp;#x014B;o" --&gt;&lt;/phoneme&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S4.6">4.6 <code>&lt;phoneme&gt;</code> Element</h3>
<p>A <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> <em title=
"MAY in RFC 2119 context" class="RFC2119">MAY</em> contain one or
more <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> elements.
The <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element
contains text describing how the <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> is pronounced.</p>
<p>The <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
contain 'character' child information items. The <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> element <em title=
"MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>
contain 'element' child information items from any namespace,
i.e. PLS or foreign namespace.</p>
<p>A <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element
<em title="MAY in RFC 2119 context" class="RFC2119">MAY</em> have
an <code>alphabet</code> attribute, which indicates the
pronunciation alphabet that is used for this <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> element only. See
<a href="#S4.1">Section 4.1</a> for the default pronunciation
alphabet. The legal values for the <code>alphabet</code>
attribute are described in <a href="#S2">Section 2</a>.</p>
<p>The <code>prefer</code> is an <em title=
"OPTIONAL in RFC 2119 context" class="RFC2119">OPTIONAL</em>
attribute, which indicates the pronunciation that <em title=
"MUST in RFC 2119 context" class="RFC2119">MUST</em> be used by a
<a href="#term-TTS">speech synthesis engine</a> when it is set to
<code>"true"</code>. See <a href="#S4.9">Section 4.9</a> for
required behavior when multiple pronunciations have
<code>prefer</code> set to <code>"true"</code>. The possible
values are: <code>"true"</code> or <code>"false"</code>. The
default value is <code>"false"</code>.</p>
<p>The prefer mechanism spans both the <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> and <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> elements. <a href=
"#S4.9">Section 4.9</a> describes how multiple pronunciations are
specified in PLS for <a href="#term-ASR">ASR</a> and <a href=
"#term-TTS">TTS</a>, and gives many examples in <a href=
"#S4.9.3">Section 4.9.3</a>.</p>
<h4 id="S4.6.1">Examples:</h4>
<p>More than one pronunciation per lexical entry:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;huge&lt;/grapheme&gt;
&lt;phoneme prefer="true"&gt;hjuːdʒ&lt;/phoneme&gt;
&lt;!-- IPA string is: "hju&amp;#x02D0;d&amp;#x0292;" --&gt;
&lt;phoneme&gt;juːdʒ&lt;/phoneme&gt;
&lt;!-- IPA string is: "ju&amp;#x02D0;d&amp;#x0292;" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>More than one written form and more than one
pronunciation:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;theater&lt;/grapheme&gt;
&lt;grapheme&gt;theatre&lt;/grapheme&gt;
&lt;phoneme prefer="true"&gt;ˈθɪətər&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;&amp;#x03B8;&amp;#x026A;&amp;#x0259;t&amp;#x0259;r" --&gt;
&lt;phoneme&gt;ˈθiːjətər&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;&amp;#x03B8;i&amp;#x02D0;j&amp;#x0259;t&amp;#x0259;r" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>An example of a <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> that changes the
pronunciation alphabet to a proprietary one.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;color&lt;/grapheme&gt;
&lt;phoneme&gt;ˈkʌlər&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;k&amp;#x028C;l&amp;#x0259;r" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;XYZ&lt;/grapheme&gt;
&lt;phoneme alphabet="x-example-alphabet"&gt;XYZ&lt;/phoneme&gt;
&lt;!-- The above pronunciation is given in a proprietary alphabet
called: "x-example-alphabet" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S4.7">4.7 <code>&lt;alias&gt;</code> Element</h3>
<p>A <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
<em title="MAY in RFC 2119 context" class="RFC2119">MAY</em>
contain one or more <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> elements which are used to
indicate the pronunciation of an <a href=
"#term-acronym">acronym</a> or an abbreviated term, in terms of
other <a href="#term-Orthography">orthographies</a>, or other
substitutions as necessary; see examples below and in <a href=
"#S4.9.3">Section 4.9.3</a>.</p>
<p>The <a href="#S4.7"><code>&lt;alias&gt;</code></a> element
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
contain 'character' child information items. The <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> element <em title=
"MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>
contain 'element' child information items from any namespace,
i.e. PLS or foreign namespace.</p>
<p>In a <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element,
both <a href="#S4.7"><code>&lt;alias&gt;</code></a> elements and
<a href="#S4.6"><code>&lt;phoneme&gt;</code></a> elements
<em title="MAY in RFC 2119 context" class="RFC2119">MAY</em> be
present. If authors want explicit control over the pronunciation,
they can use the <a href="#S4.6"><code>&lt;phoneme&gt;</code></a>
element instead of the <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> element.</p>
<p>The <a href="#S4.7"><code>&lt;alias&gt;</code></a> element has
an <em title="OPTIONAL in RFC 2119 context" class=
"RFC2119">OPTIONAL</em> <code>prefer</code> attribute analogous
to the <code>prefer</code> attribute for the <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> element; see <a href=
"#S4.6">Section 4.6</a> for a normative description of the
<code>prefer</code> attribute.</p>
<p>Pronunciations of <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> element contents
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em> be
generated by the processor using pronunciations described by the
<a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element of any
constituent graphemes in the PLS document and without invoking
recursive access to the PLS document on the <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> elements of any
constituent graphemes. The processor <em title=
"SHOULD in RFC 2119 context" class="RFC2119">SHOULD</em>
determine the pronunciations of the remaining <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> element contents by the
same process that it determines the pronunciation of
out-of-lexicon graphemes.</p>
<h4 id="S4.7.1">Examples:</h4>
<p><a href="#term-acronym-exp">Acronym expansion</a> using the
<a href="#S4.7"><code>&lt;alias&gt;</code></a> element:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;W3C&lt;/grapheme&gt;
&lt;alias&gt;World Wide Web Consortium&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>The following example illustrates a combination of <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> and <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> elements. The indicated
acronym, "GNU", has only two pronunciations. Note that the
pronunciation described by the <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> element of "Unix" is not
used as part of the pronunciation of the <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> element contents of "GNU"
as recursion of <a href="#S4.7"><code>&lt;alias&gt;</code></a> is
not permissible. The pronunciations described by the <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> elements of "GNU" and
"Unix" are used by the processor to generate the pronunciation of
"GNU is Not Unix".</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;GNU&lt;/grapheme&gt;
&lt;alias&gt;&lt;!-- be careful about recursion here --&gt;GNU is Not Unix&lt;/alias&gt;
&lt;phoneme&gt;gəˈnuː&lt;/phoneme&gt;
&lt;!-- IPA string is: "g&amp;#x0259;&amp;#x02C8;nu&amp;#x02D0;" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Unix&lt;/grapheme&gt;
&lt;grapheme&gt;UNIX&lt;/grapheme&gt;
&lt;alias&gt;a multiplexed information and computing service&lt;/alias&gt;
&lt;phoneme&gt;ˈjuːnɪks&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;ju&amp;#x02D0;n&amp;#x026A;ks" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S4.8">4.8 <code>&lt;example&gt;</code> Element</h3>
<p>The <a href="#S4.8"><code>&lt;example&gt;</code></a> element
includes an example sentence that illustrates an occurrence of
this <a href="#term-Lexeme">lexeme</a>. Because the examples are
explicitly marked, automated tools can be used for regression
testing and for generation of <a href=
"#term-Pron-Lexicon">pronunciation lexicon</a> documentation.</p>
<p>The <a href="#S4.8"><code>&lt;example&gt;</code></a> element
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
contain 'character' child information items. The <a href=
"#S4.8"><code>&lt;example&gt;</code></a> element <em title=
"MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>
contain 'element' child information items from any namespace,
i.e. PLS or foreign namespace.</p>
<p>Zero, one or many <a href=
"#S4.8"><code>&lt;example&gt;</code></a> elements <em title=
"MAY in RFC 2119 context" class="RFC2119">MAY</em> be provided
for a single <a href="#S4.4"><code>&lt;lexeme&gt;</code></a>
element.</p>
<h4 id="S4.8.1">Example:</h4>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;phoneme&gt;led&lt;/phoneme&gt;
&lt;example&gt;My feet were as heavy as lead.&lt;!-- possible comment --&gt;&lt;/example&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;phoneme&gt;liːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
&lt;example&gt;The guide once again took the lead.&lt;/example&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S4.9">4.9 Multiple Pronunciations for ASR and TTS</h3>
<p>This section describes the treatment of multiple
pronunciations specified in a PLS document for <a href=
"#term-ASR">ASR</a> and <a href="#term-TTS">TTS</a>.</p>
<h3 id="S4.9.1">4.9.1 Multiple Pronunciations for ASR</h3>
<p>If more than one pronunciation for a given <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> is specified (either by
<a href="#S4.6"><code>&lt;phoneme&gt;</code></a> elements or
<a href="#S4.7"><code>&lt;alias&gt;</code></a> elements or a
combination of both), an <a href="#term-ASR">ASR</a> processor
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
consider each of them as valid pronunciations for the <a href=
"#S4.5"><code>&lt;grapheme&gt;</code></a>. See <a href=
"#S4.9.3ex2">Example 2</a> and following examples in <a href=
"#S4.9.3">Section 4.9.3</a>.</p>
<p>If more than one <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> contains the same
<a href="#S4.5"><code>&lt;grapheme&gt;</code></a>, all relevant
pronunciations (see discussion in <a href="#S4.4">Section 4.4</a>
regarding the selection of relevant pronunciations using the
<code>role</code> attribute) will be collected in document order
and an <a href="#term-ASR">ASR</a> processor <em title=
"MUST in RFC 2119 context" class="RFC2119">MUST</em> consider all
of them as valid pronunciations for the <a href=
"#S4.5"><code>&lt;grapheme&gt;</code></a>. See <a href=
"#S4.9.3ex7">Example 7</a> and <a href="#S4.9.3ex8">Example 8</a>
in <a href="#S4.9.3">Section 4.9.3</a>.</p>
<h3 id="S4.9.2">4.9.2 Multiple Pronunciations for TTS</h3>
<p>If more than one pronunciation for a given <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> is specified (either by
<a href="#S4.6"><code>&lt;phoneme&gt;</code></a> elements or
<a href="#S4.7"><code>&lt;alias&gt;</code></a> elements or a
combination of both), a <a href="#term-TTS">TTS</a> processor
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
use the first one in document order that has the
<code>prefer</code> attribute set to <code>"true"</code>. If none
of the pronunciations has <code>prefer</code> set to
<code>"true"</code>, the <a href="#term-TTS">TTS</a> processor
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
use the first one in document order unless the <a href=
"#term-TTS">TTS</a> processor is documented as having a method of
selecting pronunciations, in which case the processor <em title=
"MUST in RFC 2119 context" class="RFC2119">MUST</em> use any one
of the pronunciations. See <a href="#S4.9.3ex2">Example 2</a> and
following examples in <a href="#S4.9.3">Section 4.9.3</a>.</p>
<p>If more than one <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> contains the same
<a href="#S4.5"><code>&lt;grapheme&gt;</code></a>, all relevant
pronunciations (see discussion in <a href="#S4.4">Section 4.4</a>
regarding the selection of relevant pronunciations using the
<code>role</code> attribute) will be collected in document order
and a <a href="#term-TTS">TTS</a> processor <em title=
"MUST in RFC 2119 context" class="RFC2119">MUST</em> use the
first one in document order that has the <code>prefer</code>
attribute set to <code>"true"</code>. If none of the relevant
pronunciations has <code>prefer</code> set to
<code>"true"</code>, the <a href="#term-TTS">TTS</a> processor
<em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
use the first one in document order unless the <a href=
"#term-TTS">TTS</a> processor is documented as having a method of
selecting pronunciations, in which case the processor <em title=
"MUST in RFC 2119 context" class="RFC2119">MUST</em> use any one
of the relevant pronunciations. See <a href="#S4.9.3ex7">Example
7</a> and <a href="#S4.9.3ex8">Example 8</a> in <a href=
"#S4.9.3">Section 4.9.3</a>.</p>
<p>Note that a <a href="#term-TTS">TTS</a> processor may have
language-dependent internal mechanisms enabling it to
automatically choose between multiple pronunciations. See
<a href="#S4.9.3ex9">Example 9</a> in <a href="#S4.9.3">Section
4.9.3</a>.</p>
<h3 id="S4.9.3">4.9.3 Examples of Multiple Pronunciations</h3>
<p><i>This section is informative.</i></p>
<p>The following examples are designed to describe and illustrate
the most common examples of multiple pronunciations. Both
<a href="#term-ASR">ASR</a> and <a href="#term-TTS">TTS</a>
behavior is described.</p>
<h4 id="S4.9.3ex1">Example 1:</h4>
<p>In the following example, there is only one pronunciation. It
will be used by both <a href="#term-ASR">ASR</a> and <a href=
"#term-TTS">TTS</a> processors.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;bead&lt;/grapheme&gt;
&lt;phoneme&gt;biːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "bi&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h4 id="S4.9.3ex2">Example 2:</h4>
<p>In the following example, there are two pronunciations. An
<a href="#term-ASR">ASR</a> processor will recognize both
pronunciations, whereas a <a href="#term-TTS">TTS</a> processor
will only use one. Since none of the pronunciations has
<code>prefer</code> set to <code>"true"</code>, unless the
processor is documented to have a different strategy, it will use
the first of the pronunciations because it is first in document
order.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;read&lt;/grapheme&gt;
&lt;phoneme&gt;red&lt;/phoneme&gt;
&lt;phoneme&gt;riːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "ri&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h4 id="S4.9.3ex3">Example 3:</h4>
<p>In the following example, there are two pronunciations. An
<a href="#term-ASR">ASR</a> processor will recognize both
pronunciations, whereas a <a href="#term-TTS">TTS</a> processor
will only use the second one (because it has <code>prefer</code>
set to <code>"true"</code>).</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;phoneme&gt;led&lt;/phoneme&gt;
&lt;phoneme prefer="true"&gt;liːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h4 id="S4.9.3ex4">Example 4:</h4>
<p>In the following example, "read" has two pronunciations. The
first one is specified by means of an alias to "red", which is
defined just below it. An <a href="#term-ASR">ASR</a> processor
will recognize both pronunciations, whereas a <a href=
"#term-TTS">TTS</a> processor will only use one. Since none of
the pronunciations has <code>prefer</code> set to
<code>"true"</code>, unless the processor is documented to have a
different strategy, it will use the first of the pronunciations
because it is first in document order. In this example, the alias
refers to a lexeme later in the lexicon, but in general, this
order is not relevant.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;read&lt;/grapheme&gt;
&lt;alias&gt;red&lt;/alias&gt;
&lt;phoneme&gt;riːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "ri&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;red&lt;/grapheme&gt;
&lt;phoneme&gt;red&lt;/phoneme&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h4 id="S4.9.3ex5">Example 5:</h4>
<p>In the following example, there are two pronunciations for
"lead". Both are given with <code>prefer</code> set to
<code>"true"</code>. An <a href="#term-ASR">ASR</a> processor
will recognize both pronunciations, whereas a <a href=
"#term-TTS">TTS</a> processor will only use the first one
(because it is first in document order that has
<code>prefer</code> set to <code>"true"</code>).</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;alias prefer="true"&gt;led&lt;/alias&gt;
&lt;phoneme prefer="true"&gt;liːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;led&lt;/grapheme&gt;
&lt;phoneme&gt;led&lt;/phoneme&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h4 id="S4.9.3ex6">Example 6:</h4>
<p>In the following example, there are two pronunciations for
"lead". <a href="#term-ASR">ASR</a> processor will recognize both
pronunciations, whereas a <a href="#term-TTS">TTS</a> processor
will only use the second one (because it has <code>prefer</code>
set to <code>"true"</code>). Note that the alias entry for "lead"
as "led" does not inherit the preference of the pronunciation of
the alias.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;alias&gt;led&lt;/alias&gt;
&lt;phoneme prefer="true"&gt;liːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;led&lt;/grapheme&gt;
&lt;phoneme prefer="true"&gt;led&lt;/phoneme&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h4 id="S4.9.3ex7">Example 7:</h4>
<p>In the following example, "lead" has two different entries in
the lexicon. An <a href="#term-ASR">ASR</a> processor will
recognize both pronunciations given here, but a <a href=
"#term-TTS">TTS</a> processor will only recognize one. Since none
of the pronunciations has <code>prefer</code> set to
<code>"true"</code>, unless the processor is documented to have a
different strategy, it will use the "led" pronunciation because
it is first in document order.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;phoneme&gt;led&lt;/phoneme&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;phoneme&gt;liːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h4 id="S4.9.3ex8">Example 8:</h4>
<p>In the following example, there are two pronunciations in each
of two different lexeme entries in the same lexicon document. An
<a href="#term-ASR">ASR</a> processor will recognize both
pronunciations given here, but a <a href="#term-TTS">TTS</a>
processor will only recognize the "liːd" pronunciation, because
it is the first one in document order that has
<code>prefer</code> set to <code>"true"</code>.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;alias&gt;led&lt;/alias&gt;
&lt;phoneme prefer="true"&gt;liːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;lead&lt;/grapheme&gt;
&lt;phoneme prefer="true"&gt;led&lt;/phoneme&gt;
&lt;phoneme&gt;liːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h4 id="S4.9.3ex9">Example 9:</h4>
<p>In the following example in French, "1" has three
pronunciations. The latter two pronunciations are specified by
means of an alias to "une", which is defined just below it. An
<a href="#term-ASR">ASR</a> processor will recognize all three
pronunciations given here, but a <a href="#term-TTS">TTS</a>
processor will only recognize the "un" pronunciation, unless
otherwise documented by the processor. A <a href=
"#term-TTS">TTS</a> processor documented capable of automatically
choosing between multiple pronunciations will select either the
"un" or "une" alias (given a grammatical context). If it selects
the "une" alias then the "yn" pronunciation will be used because
it has <code>prefer</code> set to <code>"true"</code>.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="fr"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;1&lt;/grapheme&gt;
&lt;alias&gt;un&lt;/alias&gt;
&lt;alias&gt;une&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;une&lt;/grapheme&gt;
&lt;phoneme prefer="true"&gt;yn&lt;/phoneme&gt;
&lt;phoneme&gt;ynə&lt;/phoneme&gt;
&lt;!-- IPA string is: "yn&amp;#x0259;" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h2 id="S5">5. Examples</h2>
<p><i>This section is informative.</i></p>
<h3 id="S5.1">5.1 Simple Case</h3>
<p>In its simplest form the <a href=
"#term-Pron-Lexicon">Pronunciation Lexicon</a> language allows
<a href="#term-Orthography">orthographies</a> (the textual
representation) to be associated with pronunciations (the
<a href="#term-Phoneme">phonetic/phonemic representation</a>). A
<a href="#term-Pron-Lexicon">Pronunciation Lexicon</a> document
typically contains multiple entries. So, for example, to specify
the pronunciation for proper names, such as "Newton" and
"Scahill", the markup will look like the following.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-GB"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Newton&lt;/grapheme&gt;
&lt;phoneme&gt;ˈnjuːtən&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;nju&amp;#x02D0;t&amp;#x0259;n" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Scahill&lt;/grapheme&gt;
&lt;phoneme&gt;ˈskɑhɪl&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;sk&amp;#x0251;h&amp;#x026A;l" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>Here we see the root element <a href=
"#S4.1"><code>&lt;lexicon&gt;</code></a> which contains the two
<a href="#term-Lexeme">lexemes</a> for the words "Newton" and
"Scahill". Each <a href="#S4.4"><code>&lt;lexeme&gt;</code></a>
is a composite element consisting of the <a href=
"#term-Orthography">orthographic</a> and pronunciation
representations for the entry. For each of the two <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> elements there is a
single <a href="#S4.5"><code>&lt;grapheme&gt;</code></a> element
which includes the <a href="#term-Orthography">orthographic</a>
text and a single <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> element which includes
the pronunciation. In this case the <code>alphabet</code>
attribute of the <a href="#S4.1"><code>&lt;lexicon&gt;</code></a>
element is set to <code>"ipa"</code>, so the <a href=
"#term-IPA">International Phonetic Alphabet</a> [<a href=
"#ref-ipa">IPA</a>] is being used for all the pronunciations.</p>
<h3 id="S5.2">5.2 Multiple pronunciations for the same
orthography</h3>
<p>For <a href="#term-ASR">ASR</a> systems it is common to rely
on multiple pronunciations of the same word or phrase in order to
cope with variations of pronunciation within a language. In the
<a href="#term-Lexicon">Pronunciation Lexicon</a> language,
multiple pronunciations are represented by more than one <a href=
"#S4.6"><code>&lt;phoneme&gt;</code></a> (or <a href=
"#S4.7"><code>&lt;alias&gt;</code></a>) element within the same
<a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element.</p>
<p>In the following example the word "Newton" has two possible
pronunciations.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-GB"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Newton&lt;/grapheme&gt;
&lt;phoneme&gt;ˈnjuːtən&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;nju&amp;#x02D0;t&amp;#x0259;n" --&gt;
&lt;phoneme&gt;ˈnuːtən&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;nu&amp;#x02D0;t&amp;#x0259;n" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>In the situation where only a single pronunciation needs to be
selected among multiple pronunciations that are available (for
example where a <a href="#term-Pron-Lexicon">pronunciation
lexicon</a> is being used by a <a href="#term-TTS">speech
synthesis</a> system), then the <code>prefer</code> attribute on
the <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element may
be used to indicate the preferred pronunciation.</p>
<h3 id="S5.3">5.3 Multiple orthographies</h3>
<p>In some situations there are alternative textual
representations for the same word or phrase. This can arise due
to a number of reasons. See <a href="#S4.5">Section 4.5</a> for
details. Because these are representations that have the same
meaning (as opposed to <a href="#term-Homophone">homophones</a>),
it is recommended that they be represented using a single
<a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element that
contains multiple graphemes.</p>
<p>Here are two simple examples of multiple <a href=
"#term-Orthography">orthographies</a>: alternative spelling of an
English word and multiple writings of a Japanese word.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;!-- English entry showing how alternative spellings are handled --&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;colour&lt;/grapheme&gt;
&lt;grapheme&gt;color&lt;/grapheme&gt;
&lt;phoneme&gt;ˈkʌlər&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;k&amp;#x028C;l&amp;#x0259;r" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="jp"&gt;
&lt;!-- Japanese entry showing how multiple writing systems are handled
romaji, kanji and hiragana orthographies --&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;nihongo&lt;/grapheme&gt;
&lt;grapheme&gt;日本語&lt;/grapheme&gt;
&lt;grapheme&gt;にほんご&lt;/grapheme&gt;
&lt;phoneme&gt;ɲihoŋo&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x0272;iho&amp;#x014B;o" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>In some cases the pronunciations may overlap rather than being
exactly the same. For example the English names "Smyth" and
"Smith" share one pronunciation, but "Smyth" has a pronunciation
that is only relevant to itself. Hence this needs to be
represented using multiple <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> elements.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Smyth&lt;/grapheme&gt;
&lt;grapheme&gt;Smith&lt;/grapheme&gt;
&lt;phoneme&gt;smɪθ/phoneme&gt;
&lt;!-- IPA string is: "sm&amp;#x026A;&amp;#x03B8;" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Smyth&lt;/grapheme&gt;
&lt;phoneme&gt;smaɪð&lt;/phoneme&gt;
&lt;!-- IPA string is: "sma&amp;#x026A;&amp;#x00F0;" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S5.4">5.4 Homophones</h3>
<p>Most languages have <a href="#term-Homophone">homophones</a>,
words with the same pronunciation but different meanings (and
possibly different spellings), for instance "seed" and "cede". It
is recommended that these be represented as different
lexemes.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;cede&lt;/grapheme&gt;
&lt;phoneme&gt;siːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "si&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;seed&lt;/grapheme&gt;
&lt;phoneme&gt;siːd&lt;/phoneme&gt;
&lt;!-- IPA string is: "si&amp;#x02D0;d" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S5.5">5.5 Homographs</h3>
<p>Most languages have words with different meanings but the same
spelling (and sometimes different pronunciations), called
<a href="#term-Homograph">homographs</a>. For example, in English
the word <em>bass</em> (fish) and the word <em>bass</em> (in
music) have identical spellings but different meanings and
pronunciations. Although it is recommended that these words be
represented using separate <a href=
"#S4.4"><code>&lt;lexeme&gt;</code></a> elements that are
distinguished by different values of the <code>role</code>
attribute (see <a href="#S4.4">Section 4.4</a>), if a <a href=
"#term-Pron-Lexicon">pronunciation lexicon</a> author does not
want to distinguish between the two words they could simply be
represented as alternative pronunciations within the same
<a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element. In the
latter case the <a href="#term-TTS">TTS</a> processor will not be
able to distinguish when to apply the first or the second
transcription.</p>
<p>In this example the pronunciations of the <a href=
"#term-Homograph">homograph</a> "bass" are shown.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;bass&lt;/grapheme&gt;
&lt;phoneme&gt;bæs&lt;/phoneme&gt;
&lt;!-- IPA string is: b&amp;#x00E6;s --&gt;
&lt;phoneme&gt;beɪs&lt;/phoneme&gt;
&lt;!-- IPA string is: be&amp;#x026A;s --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>Note that English contains numerous examples of noun-verb
pairs that can be treated either as <a href=
"#term-Homograph">homographs</a> or as alternative
pronunciations, depending on author preference. Two examples are
the noun/verb "refuse" and the noun/verb "address".</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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"
xmlns:mypos="http://www.example.com/my_pos_namespace"
alphabet="ipa" xml:lang="en-US"&gt;
&lt;lexeme role="mypos:verb"&gt;
&lt;grapheme&gt;refuse&lt;/grapheme&gt;
&lt;phoneme&gt;rɪˈfjuːz&lt;/phoneme&gt;
&lt;!-- IPA string is: "r&amp;#x026A;&amp;#x02C8;fju&amp;#x02D0;z" --&gt;
&lt;/lexeme&gt;
&lt;lexeme role="mypos:noun"&gt;
&lt;grapheme&gt;refuse&lt;/grapheme&gt;
&lt;phoneme&gt;ˈrefjuːs&lt;/phoneme&gt;
&lt;!-- IPA string is: "&amp;#x02C8;refju&amp;#x02D0;s" --&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h3 id="S5.6">5.6 Pronunciation by Orthography (Acronyms,
Abbreviations, etc.)</h3>
<p>For some words and phrases pronunciation can be expressed
quickly and conveniently as a sequence of other <a href=
"#term-Orthography">orthographies</a>. The developer is not
required to have linguistic knowledge, but instead makes use of
the pronunciations that are already expected to be available. To
express pronunciations using other <a href=
"#term-Orthography">orthographies</a> the <a href=
"#S4.7"><code>&lt;alias&gt;</code></a> element may be used.</p>
<p>This feature may be very useful to deal with <a href=
"#term-acronym-exp">acronym expansion</a>.</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;!--
Acronym expansion
--&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;W3C&lt;/grapheme&gt;
&lt;alias&gt;World Wide Web Consortium&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;!--
number representation
--&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;101&lt;/grapheme&gt;
&lt;alias&gt;one hundred and one&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;!--
crude pronunciation mechanism
--&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;Thailand&lt;/grapheme&gt;
&lt;alias&gt;tie land&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;!--
crude pronunciation mechanism and acronym expansion
--&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;BBC 1&lt;/grapheme&gt;
&lt;alias&gt;be be sea one&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<h2 id="S6">6. References</h2>
<h3 id="S6.1">6.1 Normative References</h3>
<dl>
<dt><a id="ref-bcp47" name="ref-bcp47">[BCP47]</a></dt>
<dd><cite><a href="http://www.rfc-editor.org/rfc/bcp/bcp47.txt"
shape="rect">Tags for the Identification of
Languages</a></cite>, A. Phillips and M. Davis, Editors. IETF,
September 2006. This RFC is available at <a href=
"http://www.rfc-editor.org/rfc/bcp/bcp47.txt" shape=
"rect">http://www.rfc-editor.org/rfc/bcp/bcp47.txt</a>.</dd>
<dt><a id="ref-ipahndbk" name=
"ref-ipahndbk">[IPAHNDBK]</a></dt>
<dd><cite><a href=
"http://www.arts.gla.ac.uk/IPA/handbook.html">Handbook of the
International Phonetic Association</a></cite>, International
Phonetic Association, Editors. Cambridge University Press, July
1999. Information on the Handbook is available at
<a href="http://www.arts.gla.ac.uk/IPA/handbook.html">
http://www.arts.gla.ac.uk/IPA/handbook.html</a>.</dd>
<dt><a id="ref-rfc2119" name="ref-rfc2119" shape=
"rect">[RFC2119]</a></dt>
<dd><cite><a href="http://www.ietf.org/rfc/rfc2119.txt" shape=
"rect">Key words for use in RFCs to Indicate Requirement
Levels</a></cite>, S. Bradner, Editor. IETF, March 1997. This
RFC is available at <a href=
"http://www.ietf.org/rfc/rfc2119.txt" shape=
"rect">http://www.ietf.org/rfc/rfc2119.txt</a>.</dd>
<dt><a id="ref-rfc3986" name="ref-rfc3986" shape=
"rect">[RFC3986]</a></dt>
<dd><cite><a href="http://www.ietf.org/rfc/rfc3986.txt" shape=
"rect">Uniform Resource Identifier (URI): Generic
Syntax</a></cite>, T. Berners-Lee et al., Editors. IETF,
January 2005. This RFC is available at <a href=
"http://www.ietf.org/rfc/rfc3986.txt" shape=
"rect">http://www.ietf.org/rfc/rfc3986.txt</a>.</dd>
<dt><a id="ref-rfc3987" name="ref-rfc3987" shape=
"rect">[RFC3987]</a></dt>
<dd><cite><a href="http://www.ietf.org/rfc/rfc3987.txt" shape=
"rect">Internationalized Resource Identifiers
(IRIs)</a></cite>, M. Duerst and M. Suignard, Editors. IETF,
January 2005. This RFC is available at <a href=
"http://www.ietf.org/rfc/rfc3987.txt" shape=
"rect">http://www.ietf.org/rfc/rfc3987.txt</a>.</dd>
<dt><a id="ref-rfc4267" name="ref-rfc4267" shape=
"rect">[RFC4267]</a></dt>
<dd><cite><a href="http://www.ietf.org/rfc/rfc4267.txt" shape=
"rect">The W3C Speech Interface Framework Media Types:
application/voicexml+xml, application/ssml+xml,
application/srgs, application/srgs+xml, application/ccxml+xml,
and application/pls+xml</a></cite>, Max Froumentin, Editor.
IETF, November 2005. This RFC is available at <a href=
"http://www.ietf.org/rfc/rfc4267.txt" shape=
"rect">http://www.ietf.org/rfc/rfc4267.txt</a>.</dd>
<dt><a id="ref-SRGS" name="ref-SRGS" shape=
"rect">[SRGS]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-speech-grammar-20040316/" shape=
"rect">Speech Recognition Grammar Specification Version
1.0</a></cite>, Andrew Hunt and Scott McGlashan, Editors. World
Wide Web Consortium, 16 March 2004. This version of the SRGS
1.0 Recommendation is <a href=
"http://www.w3.org/TR/2004/REC-speech-grammar-20040316/" shape=
"rect">http://www.w3.org/TR/2004/REC-speech-grammar-20040316/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/speech-grammar/" shape=
"rect">http://www.w3.org/TR/speech-grammar/</a>.</dd>
<dt><a id="ref-SSML" name="ref-SSML" shape=
"rect">[SSML]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/"
shape="rect">Speech Synthesis Markup Language (SSML) Version
1.0</a></cite>, Daniel C. Burnett, et al., Editors. World Wide
Web Consortium, 7 September 2004. This version of the SSML 1.0
Recommendation is <a href=
"http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/"
shape=
"rect">http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/speech-synthesis/" shape=
"rect">http://www.w3.org/TR/speech-synthesis/</a>.</dd>
<dt><a id="ref-xml-10" name="ref-xml-10" shape=
"rect">[XML10]</a></dt>
<dd><cite><a href="http://www.w3.org/TR/2006/REC-xml-20060816/"
shape="rect">Extensible Markup Language (XML) 1.0 (Fourth
Edition)</a></cite>, T. Bray et al. World Wide Web Consortium,
16 August 2006, edited in place 29 September 2006. This version
of the XML 1.0 Recommendation is <a href=
"http://www.w3.org/TR/2006/REC-xml-20060816/" shape=
"rect">http://www.w3.org/TR/2006/REC-xml-20060816/</a>. The
latest version is available at <a href=
"http://www.w3.org/TR/REC-xml/" shape=
"rect">http://www.w3.org/TR/REC-xml/</a>.</dd>
<dt><a id="ref-xml-11" name="ref-xml-11" shape=
"rect">[XML11]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2006/REC-xml11-20060816/" shape=
"rect">Extensible Markup Language (XML) 1.1 (Second
Edition)</a></cite>, T. Bray et al. World Wide Web Consortium,
16 August 2006, edited in place 29 September 2006. This version
of the XML 1.1 Recommendation is <a href=
"http://www.w3.org/TR/2006/REC-xml11-20060816/" shape=
"rect">http://www.w3.org/TR/2006/REC-xml11-20060816/</a>. The
latest version is available at <a href=
"http://www.w3.org/TR/xml11/" shape=
"rect">http://www.w3.org/TR/xml11/</a>.</dd>
<dt><a id="ref-xml-base" name="ref-xml-base" shape=
"rect">[XML-BASE]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2001/REC-xmlbase-20010627/" shape=
"rect">XML Base</a></cite>, J. Marsh, editor. World Wide Web
Consortium, 27 June 2001. This version of the XML Base
Recommendation is <a href=
"http://www.w3.org/TR/2001/REC-xmlbase-20010627/" shape=
"rect">http://www.w3.org/TR/2001/REC-xmlbase-20010627/</a>. The
latest version is available at <a href=
"http://www.w3.org/TR/xmlbase/" shape=
"rect">http://www.w3.org/TR/xmlbase/</a>.</dd>
<dt><a id="ref-xml-id" name="ref-xml-id" shape=
"rect">[XML-ID]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2005/REC-xml-id-20050909/" shape=
"rect">xml:id Version 1.0</a></cite>, J. Marsh, D. Veillard, N.
Walsh. World Wide Web Consortium, 9 September 2005. This
version of the xml:id Recommendation is <a href=
"http://www.w3.org/TR/2005/REC-xml-id-20050909/" shape=
"rect">http://www.w3.org/TR/2005/REC-xml-id-20050909/</a>. The
latest version is available at <a href=
"http://www.w3.org/TR/xml-id/" shape=
"rect">http://www.w3.org/TR/xml-id/</a>.</dd>
<dt><a id="ref-xmlns-10" name="ref-xmlns-10" shape=
"rect">[XML-NS10]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2006/REC-xml-names-20060816/" shape=
"rect">Namespaces in XML 1.0 (Second Edition)</a></cite>, T.
Bray et al., Editors. World Wide Web Consortium, 16 August
2006. This version of the XML Namespaces Recommendation is
<a href="http://www.w3.org/TR/2006/REC-xml-names-20060816/"
shape=
"rect">http://www.w3.org/TR/2006/REC-xml-names-20060816/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/xml-names/" shape=
"rect">http://www.w3.org/TR/xml-names/</a>.</dd>
<dt><a id="ref-xmlns-11" name="ref-xmlns-11" shape=
"rect">[XML-NS11]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-xml-names11-20040204/" shape=
"rect">Namespaces in XML 1.1</a></cite>, T. Bray et al.,
Editors. World Wide Web Consortium, 4 February 2004. This
version of the XML Namespaces Recommendation is <a href=
"http://www.w3.org/TR/2004/REC-xml-names11-20040204/" shape=
"rect">http://www.w3.org/TR/2004/REC-xml-names11-20040204/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/xml-names11/" shape=
"rect">http://www.w3.org/TR/xml-names11/</a>.</dd>
<dt><a id="ref-xmlschema-1" name="ref-xmlschema-1" shape=
"rect">[XML-SCHEMA-1]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/" shape=
"rect">XML Schema Part 1: Structures Second Edition</a></cite>,
H. S. Thompson, et al., Editors. World Wide Web Consortium, 28
October 2004. This version of the XML Schema Part 1
Recommendation is <a href=
"http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/" shape=
"rect">http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/xmlschema-1/" shape=
"rect">http://www.w3.org/TR/xmlschema-1/</a>.</dd>
<dt><a id="ref-xmlschema-2" name="ref-xmlschema-2" shape=
"rect">[XML-SCHEMA-2]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-xmlschema-2-20041028/" shape=
"rect">XML Schema Part 2: Datatypes Second Edition</a></cite>,
Paul V. Biron and Ashok Malhotra, Editors. World Wide Web
Consortium, 28 October 2004. This version of the XML Schema
Part 2 Recommendation is <a href=
"http://www.w3.org/TR/2004/REC-xmlschema-2-20041028/" shape=
"rect">http://www.w3.org/TR/2004/REC-xmlschema-2-20041028/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/xmlschema-2/" shape=
"rect">http://www.w3.org/TR/xmlschema-2/</a>.</dd>
</dl>
<h3 id="S6.2">6.2 Informative References</h3>
<dl>
<dt><a id="ref-dc" name="ref-dc">[DC]</a></dt>
<dd><cite><a href="http://dublincore.org/">Dublin Core Metadata
Initiative</a></cite>.<br />
See <a href=
"http://dublincore.org/">http://dublincore.org/</a>.</dd>
<dt><a id="ref-dc-es" name="ref-dc-es">[DC-ES]</a></dt>
<dd><cite><a href=
"http://dublincore.org/documents/dces/">Dublin Core Metadata
Element Set, Version 1.1: Reference
Description</a></cite>.<br />
See <a href=
"http://dublincore.org/documents/dces/">http://dublincore.org/documents/dces/</a>.</dd>
<dt><a id="ref-dict" name="ref-dict">[DICT]</a></dt>
<dd><cite><a href=
"http://dictionary.reference.com">Dictionary.com Unabridged (v
1.0.1)</a></cite>, Dictionary.com. Based on the Random House
Unabridged Dictionary, © Random House, Inc. 2006. Entries in
the dictionary are available at <a href=
"http://dictionary.reference.com">http://dictionary.reference.com</a>.</dd>
<dt><a id="ref-html" name="ref-html">[HTML]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/1999/REC-html401-19991224/">HTML 4.01
Specification</a></cite>, Dave Raggett, et al., Editors. World
Wide Web Consortium, 24 December 1999. This version of the HTML
4.01 Recommendation is <a href=
"http://www.w3.org/TR/1999/REC-html401-19991224/" shape=
"rect">http://www.w3.org/TR/1999/REC-html401-19991224/</a>. The
latest version of HTML is available at <a href=
"http://www.w3.org/TR/html/" shape=
"rect">http://www.w3.org/TR/html/</a>.</dd>
<dt><a id="ref-ipa" name="ref-ipa">[IPA]</a></dt>
<dd><cite><a href=
"http://www.arts.gla.ac.uk/IPA/ipa.html">International Phonetic
Association</a></cite>.<br />
See <a href=
"http://www.arts.gla.ac.uk/IPA/ipa.html">http://www.arts.gla.ac.uk/IPA/ipa.html</a>
for the organization's website.</dd>
<dt><a id="ref-ipaunicode1" name=
"ref-ipaunicode1">[IPAUNICODE1]</a></dt>
<dd><cite><a href=
"http://web.uvic.ca/ling/resources/ipa/charts/unicode_ipa-chart.htm">
The International Phonetic Alphabet</a></cite>, J. Esling. This
table of IPA characters in Unicode is available at <a href=
"http://web.uvic.ca/ling/resources/ipa/charts/unicode_ipa-chart.htm">
http://web.uvic.ca/ling/resources/ipa/charts/unicode_ipa-chart.htm</a>.</dd>
<dt><a id="ref-ipaunicode2" name=
"ref-ipaunicode2">[IPAUNICODE2]</a></dt>
<dd><cite><a href=
"http://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm">The
International Phonetic Alphabet in Unicode</a></cite>, J.
Wells. This table of Unicode values for IPA characters is
available at <a href=
"http://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm">http://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm</a>.</dd>
<dt><a id="ref-jeidaalphabet" name=
"ref-jeidaalphabet">[JEIDAALPHABET]</a></dt>
<dd><cite><a href=
"http://it.jeita.or.jp/document/publica/standard/summary/JEIDA-62-2000.pdf">
JEIDA-62-2000 Phoneme Alphabet</a></cite>. JEITA. An abstract
of this document (in Japanese) is available at <a href=
"http://it.jeita.or.jp/document/publica/standard/summary/JEIDA-62-2000.pdf">
http://it.jeita.or.jp/document/publica/standard/summary/JEIDA-62-2000.pdf</a>.</dd>
<dt><a id="ref-jeita" name="ref-jeita">[JEITA]</a></dt>
<dd><cite><a href="http://www.jeita.or.jp">Japan Electronics
and Information Technology Industries
Association</a></cite>.<br />
See <a href=
"http://www.jeita.or.jp">http://www.jeita.or.jp/</a>.</dd>
<dt><a id="ref-rdf-conc" name=
"ref-rdf-conc">[RDF-CONC]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/">Resource
Description Framework (RDF): Concepts and Abstract
Syntax</a></cite>, G. Klyne and J.J. Carroll, Editors. World
Wide Web Consortium, 10 February 2004. This version of the RDF
Concepts and Abstract Syntax is <a href=
"http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/">http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/rdf-concepts/">http://www.w3.org/TR/rdf-concepts/</a>.</dd>
<dt><a id="ref-rdf-schema" name=
"ref-rdf-schema">[RDF-SCHEMA]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-rdf-schema-20040210/">RDF
Vocabulary Description Language 1.0: RDF Schema</a></cite>, D.
Brickley and R. Guha, Editors. World Wide Web Consortium, 10
February 2004. This version of the RDF Schema Recommendation is
<a href=
"http://www.w3.org/TR/2004/REC-rdf-schema-20040210/">http://www.w3.org/TR/2004/REC-rdf-schema-20040210/</a>.
The latest version of RDF Schema is available at <a href=
"http://www.w3.org/TR/rdf-schema/">http://www.w3.org/TR/rdf-schema/</a>.</dd>
<dt><a id="ref-rdf-xml" name=
"ref-rdf-xml">[RDF-XMLSYNTAX]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/">RDF/XML
Syntax Specification</a></cite>, D. Beckett, Editor. World Wide
Web Consortium, 10 February 2004. This version of the RDF/XML
Syntax Recommendation is <a href=
"http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/">http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/</a>.
The latest version of the RDF XML Syntax is available at
<a href=
"http://www.w3.org/TR/rdf-syntax-grammar/">http://www.w3.org/TR/rdf-syntax-grammar/</a>.</dd>
<dt><a id="ref-REQS" name="ref-REQS">[REQS]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/WD-lexicon-reqs-20041029/">Pronunciation
Lexicon Specification (PLS) Version 1.0
Requirements</a></cite>, P. Baggia and F. Scahill, Editors.
World Wide Web Consortium, 29 October 2004. This document is a
work in progress. This version of the Pronunciation Lexicon
Requirements is <a href=
"http://www.w3.org/TR/2004/WD-lexicon-reqs-20041029/">http://www.w3.org/TR/2004/WD-lexicon-reqs-20041029/</a>.
The latest version of the Pronunciation Lexicon Requirements is
available at <a href=
"http://www.w3.org/TR/lexicon-reqs/">http://www.w3.org/TR/lexicon-reqs/</a>.</dd>
<dt><a id="ref-rfc2732" name="ref-rfc2732" shape=
"rect">[RFC2732]</a></dt>
<dd><cite><a href="http://www.ietf.org/rfc/rfc2732.txt" shape=
"rect">Format for Literal IPv6 Addresses in URL's</a></cite>,
R. Hinden, et al., Editors. IETF, December 1999. This RFC is
available at <a href="http://www.ietf.org/rfc/rfc2732.txt"
shape="rect">http://www.ietf.org/rfc/rfc2732.txt</a>.</dd>
<dt><a id="ref-sampa" name="ref-sampa">[SAMPA]</a></dt>
<dd><cite><a href=
"http://www.phon.ucl.ac.uk/home/sampa/home.htm">SAMPA computer
readable phonetic alphabet</a></cite>, J.C. Wells.<br />
See <a href=
"http://www.phon.ucl.ac.uk/home/sampa/home.htm">http://www.phon.ucl.ac.uk/home/sampa/home.htm</a>
for information on it.</dd>
<dt><a id="ref-SISR" name="ref-SISR" shape=
"rect">[SISR]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2006/CR-semantic-interpretation-20060111/"
shape="rect">Semantic Interpretation for Speech Recognition
(SISR) Version 1.0</a></cite>, Luc van Tichelen and Dave Burke,
Editors. World Wide Web Consortium, 5 April 2007. This version
of the SISR 1.0 Recommendation is <a href=
"http://www.w3.org/TR/2007/REC-semantic-interpretation-20070405/"
shape=
"rect">http://www.w3.org/TR/2007/REC-semantic-interpretation-20070405/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/semantic-interpretation/" shape=
"rect">http://www.w3.org/TR/semantic-interpretation/</a>.</dd>
<dt><a id="ref-SSML-11" name="ref-SSML-11" shape=
"rect">[SSML-11]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2008/WD-speech-synthesis11-20080620/"
shape="rect">Speech Synthesis Markup Language (SSML) Version
1.1</a></cite>, Daniel C. Burnett and 双志伟 (Zhi Wei Shuang),
Editors. World Wide Web Consortium, 20 June 2008. This version
of the SSML 1.1 Working Draft is <a href=
"http://www.w3.org/TR/2008/WD-speech-synthesis11-20080620/"
shape=
"rect">http://www.w3.org/TR/2008/WD-speech-synthesis11-20080620/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/speech-synthesis11/" shape=
"rect">http://www.w3.org/TR/speech-synthesis11/</a>.</dd>
<dt><a id="ref-VXML" name="ref-VXML" shape=
"rect">[VXML]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/REC-voicexml20-20040316/" shape=
"rect">Voice Extensible Markup Language (VoiceXML) Version
2.0</a></cite>, Scott McGlashan et al., Editors. World Wide Web
Consortium, 16 March 2004. This version of the VoiceXML 2.0
Recommendation is <a href=
"http://www.w3.org/TR/2004/REC-voicexml20-20040316/" shape=
"rect">http://www.w3.org/TR/2004/REC-voicexml20-20040316/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/voicexml20/" shape=
"rect">http://www.w3.org/TR/voicexml20/</a>.</dd>
<dt><a id="ref-XHTML2" name="ref-XHTML2">[XHTML2]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2004/WD-xhtml2-20040722/" shape=
"rect">XHTML 2.0</a></cite>, J. Axelsson et al., Editors. World
Wide Web Consortium, 22 July 2004. This version of the XML
XHTML 2.0 Working Draft is <a href=
"http://www.w3.org/TR/2004/WD-xhtml2-20040722/" shape=
"rect">http://www.w3.org/TR/2004/WD-xhtml2-20040722/</a>. The
latest version is available at <a href=
"http://www.w3.org/TR/xhtml2/" shape=
"rect">http://www.w3.org/TR/xhtml2/</a>.</dd>
<dt><a id="ref-XHTML-MTYPES" name=
"ref-XHTML-MTYPES">[XHTML-MTYPES]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/xhtml-media-types/xhtml-media-types.html"
shape="rect">XHTML Media Types</a></cite>, Ishikawa Masayasu,
Editor. World Wide Web Consortium, 1 August 2002. This version
of the W3C Note is <a href=
"http://www.w3.org/TR/xhtml-media-types/xhtml-media-types.html"
shape=
"rect">http://www.w3.org/TR/xhtml-media-types/xhtml-media-types.html</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/xhtml-media-types/" shape=
"rect">http://www.w3.org/TR/xhtml-media-types/</a>.</dd>
<dt><a id="ref-xpointer" name="ref-xpointer" shape=
"rect">[XPOINTER]</a></dt>
<dd><cite><a href=
"http://www.w3.org/TR/2003/REC-xptr-framework-20030325/" shape=
"rect">XPointer Framework</a></cite>, P. Grosso, E. Maler, J.
Marsh, N. Walsh. World Wide Web Consortium, 25 March 2003. This
version of the XPointer Framework Recommendation is <a href=
"http://www.w3.org/TR/2003/REC-xptr-framework-20030325/" shape=
"rect">http://www.w3.org/TR/2003/REC-xptr-framework-20030325/</a>.
The latest version is available at <a href=
"http://www.w3.org/TR/xptr-framework/" shape=
"rect">http://www.w3.org/TR/xptr-framework/</a>.</dd>
<dt><a id="ref-x-sampa" name="ref-x-sampa">[X-SAMPA]</a></dt>
<dd><cite><a href=
"http://www.phon.ucl.ac.uk/home/sampa/ipasam-x.pdf">Computer-coding
the IPA: a proposed extension of SAMPA</a></cite>, J.C. Wells,
University College London, 28 April 1995. This version is
available at <a href=
"http://www.phon.ucl.ac.uk/home/sampa/ipasam-x.pdf">http://www.phon.ucl.ac.uk/home/sampa/ipasam-x.pdf</a>.</dd>
<dt><a id="ref-web-arch" name=
"ref-web-arch">[WEB-ARCH]</a></dt>
<dd><cite><a href="http://www.w3.org/TR/webarch/">Architecture
of the World Wide Web, Volume One</a></cite>, I. Jacobs, N.
Walsh, World Wide Web Consortium, 15 December 2004. This
version of the Architecture of World Wide Web Recommendation is
<a href=
"http://www.w3.org/TR/webarch/">http://www.w3.org/TR/webarch/</a>.</dd>
</dl>
<h2 id="S7">7. Contributors and Acknowledgements</h2>
<p>The Contributors who provided ideas, comments, feedback and
implementation experience to improve this specification.
<i>(listed in alphabetical order)</i>:</p>
<dl>
<dd>Jeff Adams, Nuance</dd>
<dd>Kazuyuki Ashimura, W3C</dd>
<dd>Patrizio Bergallo, Loquendo</dd>
<dd>Ellen Eide, IBM</dd>
<dd>Max Froumentin, W3C</dd>
<dd>Richard Ishida, W3C</dd>
<dd>严峻 (Yan Jun), iFLYTEK</dd>
<dd>Matt Oshry, Microsoft</dd>
<dd>Dave Pawson, RNIB</dd>
<dd>Luc Van Tichelen, Nuance</dd>
</dl>
<p>The editor wishes to thank the following W3C groups for their
helpful comments: WAI and WAI/PF, I18N and MMI.</p>
<p>This specification was written with the help of the following
people <i>(listed in alphabetical order)</i>:</p>
<dl>
<dd>Debbie Dahl, Conversational Technologies</dd>
<dd>Ken Davies, HeyAnita</dd>
<dd>Kurt Fuqua, Vail Systems</dd>
<dd>Will Gardella, SAP</dd>
<dd>Makoto Hirota, Canon</dd>
<dd>Jim Larson, Intervoice</dd>
<dd>Dave Raggett, W3C/Volantis</dd>
<dd>Matt Womer, W3C</dd>
</dl>
<h2 id="AppA">Appendix A - Schema for Pronunciation Lexicon
Specification</h2>
<p><b><i>This section is normative.</i></b></p>
<p>There are two schemas which can be used to validate PLS
documents.<br />
The latest version of the schemas are available at:</p>
<ul>
<li>The XML schema:<br />
<code>"http://www.w3.org/TR/pronunciation-lexicon/pls.xsd"</code></li>
<li>The RELAX NG schema:<br />
<code>"http://www.w3.org/TR/pronunciation-lexicon/pls.rng"</code></li>
</ul>
<p>For stability it is <em title=
"RECOMMENDED in RFC 2119 context" class=
"RFC2119">RECOMMENDED</em> that you use the dated URI available
at:</p>
<ul>
<li>The XML schema:<br />
<code>"http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20081014/pls.xsd"</code></li>
<li>The RELAX NG schema:<br />
<code>"http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20081014/pls.rng"</code></li>
</ul>
<h2 id="AppB">Appendix B - MIME Type and File Suffix</h2>
<p><b><i>This section is normative.</i></b></p>
<p>The media type associated to Pronunciation Lexicon
Specification documents is <code>"application/pls+xml"</code> and
the filename suffix is <code>".pls"</code> as defined in
[<a href="#ref-rfc4267">RFC4267</a>].</p>
<h2 id="AppC">Appendix C - Issues in Retrieving Lexical
Content</h2>
<p><i>This section is informative.</i></p>
<p>Speech applications that use a PLS document need a mechanism
enabling them to retrieve appropriate lexical content. In the
simplest of cases, an application will search the PLS document
for <a href="#S4.5"><code>&lt;grapheme&gt;</code></a> elements
with content that exactly matches the input and retrieve all
corresponding lexemes. In general, however, the retrieval of
lexical content is not so trivial; it is necessary to define what
constitutes an exact match and which lexemes are to be retrieved
when competing matches can apply.</p>
<p>Here is an example of an approach to retrieve appropriate
lexical content.</p>
<ul>
<li>The text can be tokenized into the smallest orthographic
units that might have some meaning to the written language
addressed by the PLS document. By default, for written
languages with an alphabet, all consecutive members of the
alphabet can be grouped into a single token and all other
characters (such as punctuation, hyphen and apostrophe) can be
considered as individual tokens. Applications that address
languages using a logographic writing system might consider
each logogram as a separate token. Languages that are heavily
agglutinative need a more elaborate default tokenization
procedure.</li>
<li>A speech application may provide a mechanism to segment the
text into explicit tokens. Such a mechanism can be used to
overcome a limitation of the default tokenization. For example,
"don't" can be explicitly tokenized as "do" and "n't" in order
to match a <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
element with content "n't".</li>
<li>Tokens should be processed in the direction of normal
script reading. Thus, English is processed from left to right,
and languages such as Arabic and Hebrew, from right to
left.</li>
<li>Precedence should be given to the retrieval of lexemes
having a <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
element whose content exactly matches the longest possible
sequence of consecutive tokens. Thus, a lexeme for "they'll"
should have precedence over a lexeme for "they" given the input
"they'll'.</li>
<li>Lexical retrieval should be performed by the bias of tokens
rather than characters. Thus, a lexeme for "do" should not
match the beginning of "done".</li>
<li>A match can occur only when the Unicode characters are
equal. Thus, matches should be sensitive to case and
diacritics. "Lima" should not match "lima", "cure" should not
match "curé", and "vitæ" should not match "vitae". An
application may choose to expand ligatures (such as "æ") into
corresponding letter sequences ("ae") and to use a collation
table to further loosen the match.</li>
<li>Multiple consecutive white space characters should be
reduced to a unique single white space character. Thus, "New
York" should match "New&nbsp;&nbsp;&nbsp;&nbsp; York".</li>
</ul>
<p>This outlined approach is designed principally with the needs
of English in mind and should be modified to accommodate the
particular requirements of other languages.</p>
<p>It is recommended for applications that use a PLS document to
describe the approach they adopt in retrieving lexical
content.</p>
<h4 id="AppC.1">Example:</h4>
<p>An application that uses the following PLS document:</p>
<pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;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="ipa" xml:lang="en-US"&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;New York&lt;/grapheme&gt;
&lt;alias&gt;NY&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;lexeme&gt;
&lt;grapheme&gt;York City&lt;/grapheme&gt;
&lt;alias&gt;YC&lt;/alias&gt;
&lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
<p>should process "New&nbsp;&nbsp; York City" as "NY City" rather
than "New&nbsp;&nbsp; YC" if it uses the above approach.</p>
<h2 id="AppD">Appendix D - Changes</h2>
<p><i>This section is informative.</i></p>
<h3 id="AppD.5">Changes in this document:</h3>
<ul>
<li>Modified the third example because there was a missing "uri"
attribute of &lt;lexicon&gt; element (4.4).</li>
<li>Updated IPA references (6.2).</li>
<li>The schema files (pls.xsd and pls.rng) have been
updated to allow for non-PLS attributes on all PLS elements (App. A).</li>
</ul>
<h3 id="AppD.4">Changes in the Proposed Recommendation (18
August 2008):</h3>
<ul>
<li>Updated definition of URI (1.5).</li>
<li>Clarified usage of white space in IPA transcriptions
(2).</li>
<li>Clarified definition of xml:base attribute (4.1).</li>
<li>Updated example of alternative textual representations for
German (4.5).</li>
<li>Applied minor editorial changes.</li>
</ul>
<h3 id="AppD.3">Changes in the Candidate Recommendation (12
December 2007):</h3>
<ul>
<li>Added support of IRI and XML 1.1.</li>
<li>Expanded Conformance section (3).</li>
<li>Clarified recursion language for alias element (4.7).</li>
<li>Clarified prefer attribute definition (4.6, 4.7, 4.9).</li>
<li>Updated schemas and added dated URI for stability (App.
A).</li>
<li>Added a note on QNames (4.4).</li>
<li>Added informative notes on IPA (2).</li>
<li>Updated Authors/Contributors/Acknowledgements (7).</li>
<li>Added Appendix to list changes since previous drafts (App.
D).</li>
<li>Reviewed normative/informative attributions.</li>
<li>Added Chinese example (4.4).</li>
<li>Updated examples (i.e. added SchemaLocation, fixes to IPA
codes, etc).</li>
<li>Added clarifications and fixed spelling errors.</li>
</ul>
<h3 id="AppD.2">Changes in the second Last Call Working Draft (26
October 2006):</h3>
<ul>
<li>Added a new role attribute to the lexeme element which
helps to disambiguate Homographs (4.4).</li>
<li>Added an informative appendix on issues in retrieving
lexical content (App. C).</li>
<li>Removed orthography attribute in the grapheme element
(4.4).</li>
<li>Removed recursion of alias elements (4.7).</li>
<li>Styled normative language covered by RFC2119.</li>
<li>Updated schemas (App. A).</li>
<li>Modified examples.</li>
<li>Added requested clarifications.</li>
</ul>
<h3 id="AppD.1">Changes in the first Last Call Working Draft (31
January 2006):</h3>
<ul>
<li>Support for the IPA phonetic alphabet is mandatory
(2).</li>
<li>Added a new section on multiple pronunciations (4.9).</li>
<li>Clarified the use of the "prefer" attribute (4.6, 4.7,
4.9).</li>
<li>Added a glossary of terms (1.5).</li>
<li>Fixed XML schema and added Relax NG schema (App. A).</li>
</ul>
<!-- <p><a href="http://validator.w3.org/check?uri=referer"><img src=
"http://www.w3.org/Icons/valid-xhtml10" alt="Valid XHTML 1.0!" height="31" width="88" /></a></p><p>
<a href="http://jigsaw.w3.org/css-validator/">
<img style="border:0;width:88px;height:31px"
src="http://jigsaw.w3.org/css-validator/images/vcss"
alt="Valid CSS!" />
</a>
</p> -->
</body>
</html>