WD-voice-grammar-reqs-19991223 24.6 KB

Raw Blame History Permalink

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<html>
<head>
<meta http-equiv="Content-Type" content=
"text/html; charset=iso-8859-1">
<title>Grammar Requirements for Voice Markup</title>
<link rel="stylesheet" type="text/css" href=
"http://www.w3.org/StyleSheets/TR/W3C-WD.css">
<style type="text/css">
body {
font-family: sans-serif;
margin-left: 10%;
margin-right: 5%;
color: black;
background-color: white;
background-attachment: fixed;
background-image: url(http://www.w3.org/StyleSheets/TR/WD.gif);
background-position: top left;
background-repeat: no-repeat;
font-family: Tahoma, Verdana, "Myriad Web", Syntax, sans-serif;
}
.unfinished {  font-style: normal; background-color: #FFFF33}
.dtd-code {  font-family: monospace;
 background-color: #dfdfdf; white-space: pre;
 border: #000000; border-style: solid;
 border-top-width: 1px; border-right-width: 1px;
 border-bottom-width: 1px; border-left-width: 1px; }
p.copyright {font-size: smaller}
h2,h3 {margin-top: 1em;}
.extra { font-style: italic; color: #338033 }
code {
    color: green;
    font-family: monospace;
    font-weight: bold;
}
.example {
    border: solid green;
    border-width: 2px;
    color: green;
    font-weight: bold;
    margin-right: 5%;
    margin-left: 0;
}
.bad  {
    border: solid red;
    border-width: 2px;
    margin-left: 0;
    margin-right: 5%;
    color: rgb(192, 101, 101);
}
div.navbar { text-align: center; }
div.contents {
    background-color: rgb(204,204,255);
    padding: 0.5em;
    border: none;
    margin-right: 5%;
}
table {
    margin-left: -4%;
    margin-right: 4%;
    font-family: sans-serif;
    background: white;
    border-width: 2px;
    border-color: white;
  }
th { font-family: sans-serif; background: rgb(204, 204, 153) }
td { font-family: sans-serif; background: rgb(255, 255, 153) }
.tocline { list-style: none; }
</style>
</head>
<body>
<div class="head">
<p><a href="http://www.w3.org/"><img class="head" src=
"http://www.w3.org/Icons/WWW/w3c_home.gif" alt="W3C"></a></p>

<h1 class="head">Grammar Representation Requirements for Voice
Markup Languages</h1>

<h3 class="notoc">W3C Working Draft <i>23 December 1999</i></h3>

<dl>
<dt>This version:</dt>

<dd><a href=
"http://www.w3.org/TR/1999/WD-voice-grammar-reqs-19991223">
http://www.w3.org/TR/1999/WD-voice-grammar-reqs-19991223</a></dd>

<dt>Latest version:</dt>

<dd><a href=
"http://www.w3.org/TR/voice-grammar-reqs">
http://www.w3.org/TR/voice-grammar-reqs</a></dd>

<dt>Previous versions (Member-only):</dt>

<dd><a href=
"http://www.w3.org/Voice/Group/1999/grammar-reqs-19991116.html">
http://www.w3.org/Voice/Group/1999/grammar-reqs-19991116</a></dd>

<dt>Editor:</dt>

<dd>M. K. Brown, Bell Labs, Murray Hill, NJ</dd>
</dl>

<p class="copyright"><a href=
"http://www.w3.org/Consortium/Legal/ipr-notice#Copyright">
Copyright</a> &#169; 1999 <a href="http://www.w3.org/">
W3C</a><sup>&#174;</sup> (<a href=
"http://www.lcs.mit.edu/">MIT</a>, <a href=
"http://www.inria.fr/">INRIA</a>, <a href=
"http://www.keio.ac.jp/">Keio</a>), All Rights Reserved. <abbr
title="World Wide Web Consortium">W3C</abbr> <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>, <a href=
"http://www.w3.org/Consortium/Legal/copyright-documents">document
use</a> and <a href=
"http://www.w3.org/Consortium/Legal/copyright-software">software
licensing</a> rules apply.</p>

<hr>
</div>

<h2 class="notoc">Abstract</h2>

<p>The W3C Voice Browser working group aims to develop
specifications to enable access to the Web using spoken
interaction. This document is part of a set of requirements
studies for voice browsers, and provides details of the
requirements for grammars for speech recognition.</p>

<h2>Status of this document</h2>

<p>This document describes the requirements for grammars used for
speech recognition, as a precursor to starting work on
specifications. Related requirement drafts are linked from the <a
href="/TR/1999/WD-voice-intro-19991223">introduction</a>. The
requirements are being released as working drafts but are not
intended to become proposed recommendations.</p>

<p>This specification is a Working Draft of the Voice Browser working
group for review by W3C members and other interested parties.  This is
the first public version of this document. It is a draft document and
may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use W3C Working Drafts as reference
material or to cite them as other than "work in progress".</p>

<p>Publication as a Working Draft does not imply endorsement by
the W3C membership, nor of members of the Voice Browser working
groups. This is still a draft document and may be updated,
replaced or obsoleted by other documents at any time. It is
inappropriate to cite W3C Working Drafts as other than "work in
progress."</p>

<p>This document has been produced as part of the <a href=
"http://www.w3.org/Voice/">W3C Voice Browser Activity</a>,
following the procedures set out for the <a href=
"http://www.w3.org/Consortium/Process/">W3C Process</a>. The
authors of this document are members of the <a href=
"http://www.w3.org/Voice/Group">Voice Browser Working Group</a>.
This document is for public review. Comments should be sent to
the public mailing list &lt;<a href=
"mailto:www-voice@w3.org">www-voice@w3.org</a>&gt; (<a href=
"http://www.w3.org/Archives/Public/www-voice/">archive</a>) by
14th January 2000.</p>

<p>A list of current W3C Recommendations and other technical
documents can be found at <a href="http://www.w3.org/TR">
http://www.w3.org/TR</a>.</p>

<h2>0. Introduction</h2>

<p>The main goal of this subgroup is to define a speech
recognition grammar specification language that will be generally
useful across a variety of speech platforms used in the context
of a dialog and synthesis markup environment. The process will
consist of four main steps:<br>
</p>

<ol>
<li>establish an appropriate set of requirements for grammar
specifications</li>

<li>evaluate existing grammar languages for satisfaction of
requirements</li>

<li>settle upon a language specification or modify as
necessary</li>

<li>deliver a specific language proposal to the full W3C working
group.</li>
</ol>

<br>
<p>The scope of issues discussed includes semantics and contexts
as well as natural language syntax. Therefore the activities of
the Grammar Representation Subgroup are to be coordinated with
the activities of both the Natural Language Subgroup and the
Dialog Subgroup.</p>

<p>The following eight main topic areas have been identified as
important:<br>
</p>

<ol>
<li>Natural Language Syntax</li>

<li>Large Vocabulary/Dictation</li>

<li>Grammar Contexts</li>

<li>Semantics</li>

<li>Post-Processing Issues</li>

<li>Efficiency Issues</li>

<li>XML Compatibility</li>

<li>Grammar Specification Language Syntax</li>
</ol>

<p>Each topic area consists of several issues that will be
discussed in detail in the following sections. Example
specifications presented in this document are for illustration
purposes only and do not necessarily represent recommended
formats.</p>

<h3>0.1 Terminology</h3>

<table border="1" cellpadding="6" width="85%" summary="first
column gives priority name, second its description">
<tr>
<th>BNF</th>
<td>Backus-Naur Format.</td>
</tr>

<tr>
<th>Context</th>
<td>A <b>context</b> is a subset of the full domain. A context
can possess <b>state</b>.</td>
</tr>

<tr>
<th>CFG</th>
<td>Context-Free Grammar.</td>
</tr>

<tr>
<th>Domain</th>
<td>The scope of task semantics over which the associated <b>
language</b> and associated attribute-values are meaningful.</td>
</tr>

<tr>
<th>Grammar</th>
<td>The representation of constraints defining the set of
allowable sentences in the <b>language</b>.</td>
</tr>

<tr>
<th>Language</th>
<td>The collection or set of sentences associated with a
particular <b>domain</b>. Language may refer to natural or
program language.</td>
</tr>

<tr>
<th>N-Best</th>
<td>Top N hypotheses; from speech recognition, in this case, but
could be from natural language processing.</td>
</tr>

<tr>
<th>N-Gram</th>
<td>Probabilistic grammar using conditional probabilities
P(w<sub>n</sub> | w<sub>n-1</sub> w<sub>n-2</sub> ...).</td>
</tr>

<tr>
<th>OOV</th>
<td>Out Of Vocabulary (words).</td>
</tr>

<tr>
<th>State</th>
<td>The current condition or value of variables and attributes of
a system.</td>
</tr>

<tr>
<th>URI</th>
<td>Universal Resource Identifier.</td>
</tr>

<tr>
<th>URL</th>
<td>Universal Resource Locator.</td>
</tr>

<tr>
<th>XML</th>
<td>Extensible Markup Language.</td>
</tr>
</table>

<h3>0.2 Symbology</h3>

<h4>Regular Expressions:</h4>

<table border="1" cellpadding="6" width="85%" summary="first
column gives feature name, second its description">
<tr>
<th>?</th>
<td>Postfix operator; Zero or one occurrence</td>
</tr>

<tr>
<th>*</th>
<td>Postfix operator; Zero or more occurrences</td>
</tr>

<tr>
<th>+</th>
<td>Postfix operator; One or more occurrences</td>
</tr>

<tr>
<th>()</th>
<td>Scoping symbols</td>
</tr>

<tr>
<th>( ... | ... )</th>
<td>Disjunction; exclusive OR</td>
</tr>

<tr>
<th>A:B</th>
<td>Acceptor token; input A yields output B</td>
</tr>

<tr>
<th>.</th>
<td>&lt;FILLER&gt; or equivalent</td>
</tr>
</table>

<h4>Proposed Markup Tags (so far):</h4>

<table border="1" cellpadding="6" width="85%" summary="first
column gives element, second its description, and the third any
associated attributes">
<tr align="center">
<th><tt>Tag</tt></th>
<th>Definition</th>
<th>Attributes<br>
<i>(definitions in text)</i></th>
</tr>

<tr align="center">
<td><tt>&lt;ALPHABET= ... &gt;</tt></td>
<td>Phonetic alphabet definition.</td>
<td>Phonetic alphabets.</td>
</tr>

<tr align="center">
<td><tt>&lt;FILLER&gt;</tt></td>
<td>Generic tag for OOV word(s).</td>
<td>None.</td>
</tr>

<tr align="center">
<td><tt>&lt;GRAMMAR&gt; ... &lt;/GRAMMAR&gt;</tt></td>
<td>Grammar definition section.</td>
<td><tt>TYPE, ARY</tt></td>
</tr>

<tr align="center">
<td><tt>&lt;ITEM&gt; ... &lt;/ITEM&gt;</tt></td>
<td>XML grammar rule item.</td>
<td>None.</td>
</tr>

<tr align="center">
<td><tt>&lt;N-GRAM ... &gt; ... &lt;/N-GRAM&gt;</tt></td>
<td>N-gram token specifier.</td>
<td><tt>ARY, P, PBO</tt></td>
</tr>

<tr align="center">
<td><tt>&lt;RULE ... &gt; ... &lt;/RULE&gt;</tt></td>
<td>XML format grammar rule.</td>
<td><tt>NAME</tt></td>
</tr>
</table>

<h3>0.3 Priorities</h3>

<p>The following priorities are used to indicate the level of
importance of each requirement in this document.</p>

<table border cellspacing="1" cellpadding="6" width="85%"
summary="term in first column, explanation in second">
<tr>
<th>Must Specify</th>
<td>The specification must define the feature.</td>
</tr>

<tr>
<th>Should Specify</th>
<td>The specification should define the feature, if
possible.</td>
</tr>

<tr>
<th>nice to specify</th>
<td>The specification may optionally define the feature.</td>
</tr>

<tr>
<th>Future Revision</th>
<td>The feature needs additional study before specification.</td>
</tr>
</table>

<h3>0.4 Subgroup Coordination</h3>

<p>The requirements and specification of the Grammar
Representation Subgroup will be coordinated with overlapping
requirements and specification of the Natural Language, Dialogue,
and Universal Access subgroups.<br>
</p>

<h2>1. Natural Language Syntax</h2>

<p>Specification of the natural language syntax in the grammar
representation shall conform to the following requirements:</p>

<h4>1.1 Context-Free Grammar Support (must specify)</h4>

<p>The grammar representation must support the definition of a
Context-Free Grammar (CFG) and, by subsumption, a Finite-State
Grammar (FSG). Some platforms will not support recursive rules so
content developers will need to be aware of specific product
limitations.</p>

<h4>1.2 CFG Specification (must specify)</h4>

<p>CFG's must be represented by specification in a well known
format. Each CFG rule will be a regular expression.</p>

<p class="extra">An example well-known CFG representation is
Backus-Naur Format (BNF) and a regular expression syntax is the
well known {<i>expr</i>}?, {<i>expr</i>}*, and {<i>expr</i>}+ for
optional, zero or more, and one or more expressions,
respectively, and the use of { <i>expr</i> | <i>expr</i> } for
exclusive OR.</p>

<h4>1.3 N-Gram Grammar Enabled (should specify)</h4>

<p>The grammar representation should enable the definition of an
N-Gram Grammar (NGG).</p>

<p></p>

<p class="extra">An example format is described in the
Appendix.</p>

<h4>1.4 Out-of-Vocabulary Words (must specify)</h4>

<p>The grammar representation must support the processing of
out-of-vocabulary (OOV) words and define a method for
representing such words. The content developer can specify the
action to be taken upon encountering OOV words.</p>

<h4>1.5 Disfluency and Noise Management (should specify)</h4>

<p>The grammar representation should support the handling of
disfluency and noise.</p>

<p class="extra">An example token syntax for a speech noise
absorbing model is <tt>&lt;FILLER&gt;</tt>.</p>

<h4>1.6 Notational Convenience (nice to specify, future
study)</h4>

<p>The grammar representation might allow for certain syntactic
conveniences. An example is the permutation of items in a list of
N items taken M=N at a time.</p>

<p class="extra">An example format for permutation is:</p>

<blockquote class="extra">Symbols: A, B, C<br>
<tt>'A || B'</tt> means <tt>'AB | BA'</tt><br>
<tt>'A || B || C'</tt> means <tt>'ABC | ACB | BAC | BCA | CAB |
CBA'</tt></blockquote>

<h2>2. Large Vocabulary and Dictation</h2>

<p>Special consideration for large vocabularies shall include the
following:</p>

<h4>2.1 Large Vocabulary Definition (must specify; cf. section
1.3)</h4>

<p>The grammar representation must support the definition large
vocabularies suitable for dictation applications.&#160;
Associated attributes of such grammars shall be made available to
the speech recognizer for improving interpretation
characteristics. Specifications for large vocabulary will not
preclude the definition of small grammars.</p>

<h4>2.2 Efficiency (must specify; cf. section 6)</h4>

<p>The grammar representation must not preclude efficient
processing of speech grammars with large vocabularies.<br>
</p>

<h2>3. Grammar Contexts</h2>

<p>Multiple grammar contexts shall be supported with the
following requirements:</p>

<h4>3.1 External Grammars (must specify)</h4>

<p>The grammar representation must support the inclusion of
grammars defined outside of the current context. Access to
grammar contexts shall be provided by a suitable reference
mechanism.</p>

<h4>3.2 Run-Time Definition (must specify)</h4>

<p>Grammar constraint rules must be re-definable in part or
entirety while the system is operating. Several mechanisms are
possible including, but not limited to, unconstrained
redefinition of inferior grammar rules, prior declaration of
volatile rules, partitioning of the rule space into static and
dynamic arenas, etc.&#160;&#160;<br>
</p>

<h2>4. Semantics</h2>

<p>Semantic specifications are to be coordinated with the Natural
Language, Dialog and Universal Access subgroups. In many cases
semantic definitions required by these other groups will be
implemented as part of the specification language of associated
grammars.</p>

<h4>4.1 Semantics Support (must specify)</h4>

<p>The grammar representation must support the specification of
semantics in association with the grammar syntax.</p>

<h4>4.2 Semantic Tagging (must specify)</h4>

<p>The grammar representation must support the tagging of syntax
for semantic interpretation. Semantic values shall be returned as
attribute-values pairs.</p>

<h4>4.3 Attributes (must specify)</h4>

<p>The grammar representation must include attributes that can be
attached to data returned from the speech recognizer. Such
attributes can have multiple values, be used to indicate the
context for interpretation of recognizer output, and generally
pass semantic information to later processing stages The
specification must be consistent with the natural language
interpreter input format.</p>

<h4>4.4 Attribute Processing (must specify)</h4>

<p>A grammar referencing another grammar having attributes must
be capable of performing a [currently undefined] set of
operations upon the referenced attributes.</p>

<p></p>

<p class="extra">Examples of such processing include boolean
operations, string manipulation and attribute renaming.</p>

<h2>5. Post-Processing Issues</h2>

<h4>5.1 Confidence Scoring (must specify)</h4>

<p>The grammar representation must provide information for the
post-processing of recognition confidence scores with regard to
error rejection processing. Such information can include the
language model perplexity (high perplexity would typically reduce
confidence, and hence rejection threshold) or direct cues to
tighten or relax the normal rejection constraints to provide
content based control of performance.</p>

<h4>5.2 N-Best Hypotheses (must specify)</h4>

<p>The grammar representation must support the post-processing of
N-best output of recognition hypotheses. This requirement will be
coordinated with the Dialog subgroup.<br>
</p>

<h2>6. Efficiency Issues (cf. section 2.2)</h2>

<h4>6.1 Native Grammar Formats (must specify)</h4>

<p>The grammar representation must support the downloading of
native grammar formats for efficiency purposes. A binary
reference format can be defined for this purpose. Native formats
will be useful when content is specifically written for
particular platforms.</p>

<h4>6.2 Grammar Libraries (should specify)</h4>

<p>The grammar representation should support the use of grammar
libraries, alternatively called grammar objects, that can contain
prepackaged collections of sub-grammars to be included in higher
level grammar constructs. Such libraries will be accessible via
the naming conventions, may contain symbol tables for efficient
reference resolution, may be imported and be designated to remain
resident in a platform.<br>
</p>

<h2>7. XML Compatibility</h2>

<h4>7.1 XML Embedding (must specify)</h4>

<p>The grammar representation must support easy embedding of
grammars into XML.</p>

<h4>7.2 Pure XML Format (must specify)</h4>

<p>A pure XML format for specification of grammars, including
CFG's, must be supported. XML grammar specifications must be
capable of expressive power equivalent to BNF specifications.</p>

<p><br>
</p>

<h2>8. Grammar Specification Language Syntax</h2>

<p>This is a general requirements section into which all other
requirements will eventually migrate as the representation syntax
is defined to satisfy those requirements.</p>

<h4>8.1 Understandability (must specify)</h4>

<p>The grammar representation must be easy to understand, using
well known methods for specifying the various elements.</p>

<p class="extra">Backus-Naur Format is an example of finite-state
and context-free grammar representations.</p>

<p class="extra">A modified form of the well known MIT format is
an example format for representation of N-gram grammars (cf.
Appendix).</p>

<h4>8.2 Mixed-Mode Grammars (should specify)</h4>

<p>The grammar representation should support the simultaneous
mixing of finite-state, context-free, and N-gram grammars.</p>

<h4>8.3 Language Extensions (should specify)</h4>

<p>The grammar representation should support the extension of the
grammar representation in an obvious manner.</p>

<h4>8.4 Grammar Naming (must specify)</h4>

<p>The grammar representation must support the naming of
grammars. Reference to full grammars and rules within grammars
shall be supported by a suitable multi-part naming mechanism.
Easy name resolution and overloading shall be supported. A
namespace mechanism to avoid naming conflicts shall be supported.
Such reference shall include reference by Universal Resource
Identifier (URI). Attribute fields shall be included in the
naming format.</p>

<h4>8.5 Native Natural Language (must specify)</h4>

<p>The grammar representation must support the specification of a
native language or locale. This specification can be embedded
within a grammar rule to change the native language in
mid-sentence.</p>

<p class="extra">An example syntax for the specification of
English: <tt><b>xml:lang="en"</b></tt>.</p>

<h4>8.6 Rule Weighting (must specify)</h4>

<p>The grammar representation must support the weighting of
grammar rules in the CFG format. Weighting is implicit in the
N-Gram format.</p>

<h4>8.7 Phonetic Pronunciation (must specify)</h4>

<p>The grammar representation must support the inclusion of
phonetic pronunciation rules. This information may override
default rules defined by the speech processing platform. The
thorough definition of this subject will be the charter of
another subgroup.</p>

<p></p>

<p class="extra">An example format is:</p>

<blockquote class="extra">a tag identifying the phonetic alphabet
in use e.g.<br>
<tt>&lt;alphabet=[arpabet|sampa|vendorspecifi]&gt;</tt>.</blockquote>

<h4>8.8 Grammar File Inclusion (must specify)</h4>

<p>The grammar representation must support the inclusion of other
grammar files referenced by name via a Universal Resource
Identifier (URI). This inclusion method is distinguished from
grammar reference by symbol.</p>

<h4>8.9 Comments (must specify)</h4>

<p>The grammar representation must include a commenting
mechanism. This mechanism can be provided by HTML or XML
commenting formats.</p>

<h4>8.10 Character Encodings (must specify)</h4>

<p>The grammar representation must support the use of character
encoding for foreign language support.</p>

<p></p>

<p class="extra">Example formats can include Unicode and JIS. XML
character encoding can be used for XML grammar
specifications.</p>

<h4>8.11 Recognizer Timeout Periods (should specify)</h4>

<p>The grammar representation should support the specification of
time limits inherently related to grammar characteristics.&#160;
Such inherent characteristics can include the expected (typically
maximum) times required to normally speak a sentence from the
grammar. Such time limits can directly indicate the maximum
sentence length in the grammar and may include, but not be
limited to: maximum initial silence waiting time, minimum spoken
utterance time, maximum spoken utterance time, and maximum
intra-sentence silence time (for ASR endpointing).</p>

<h2>Appendix - Additional Examples</h2>

<h4>N-Gram Grammar Format</h4>

<p>An example format is derived from the MIT N-gram format as
follows:</p>

<p>Here is a brief description of the MIT bigram file format:</p>

<blockquote><tt>
    ...comments...<br>
    ngram 1=A<br>
    ngram 2=B<br>
    ...comments...<br>
    P(w<sub>1</sub>) w<sub>1</sub> P<sub>bo</sub>(w<sub>1</sub>)<br>
    &#160;&#160;&#160;&#160;...<br>
    ...comments...<br>
    P(w<sub>2</sub>|w<sub>1</sub>) w<sub>1</sub> w<sub>2</sub><br>
    &#160;&#160;&#160;&#160;...
</tt></blockquote>

<p>where A is the number of unigrams, B is the number of bigrams,
P(w<sub>1</sub>) is the unigram probability of word w<sub>1</sub>
(symbol), P<sub>bo</sub>(w<sub>1</sub>) is the corresponding
back-off probability, P(w<sub>2</sub>|w<sub>1</sub>) is the
probability of w<sub>2</sub> conditioned on prior w<sub>1</sub>.
Start and end of sentence are indicated by the '#' symbol, such
as P(#|w<sub>i</sub>) to indicate a sentence begins with
w<sub>i</sub> and P(w<sub>j</sub>|#) to indicate the sentence
ends with w<sub>j</sub>.</p>

<p>To adapt this to arbitrary N-grams we need to either indicate
N or define a section end marker. We can also provide data type
markers such as &lt;GRAMMAR&gt; and &lt;/GRAMMAR&gt;. How about a
start marker &lt;GRAMMAR TYPE="N-GRAM" ARY=[number]&gt; to
indicate the n-ary of the grams. Comments can follow HTML style.
An alternative to specifying 1=A, 2=B, etc. is to use markers to
identify data.&#160; Examples:</p>

<p>(Feedback: N-gram span is implicit within the format.)</p>

<pre>
    &lt;N-GRAM ARY="1" P="0.01" PBO="0.001"&gt;
           "word"
    &lt;/N-GRAM&gt;
    &lt;N-GRAM ARY="2" P="0.01"&gt;
           "word list"
    &lt;/N-GRAM&gt;
</pre>

<p>(Feedback: This may be too verbose.)</p>

<p>in which case we could intermix the types, but this would make
it more difficult for systems designers to automatically allocate
resources. Another issue is whether we want to allow the symbols
to be word phrases, so I suppose we could require quote marks
around symbol strings for this (as shown).</p>

<h2>Acknowledgments</h2>
<h3>Subgroup Members</h3>

<blockquote>Michael Brown (Bell Labs)<br>
Deborah Dahl (Unisys)<br>
Charles Hemphill (Conversa)<br>
Andrew Hunt (Sun Labs)<br>
Robert Keiller (Canon)<br>
Tetsuo Kosaka (Canon)<br>
James Larson (Intel)<br>
William Ledingham (SpeechWorks)<br>
Bruce Lucas (IBM)<br>
Jens Marschner (Philips)<br>
Scott McGlashen (PipeBeach)<br>
Michael Phillips (SpeechWorks)<br>
Stephen Potter (Entropic)<br>
David Raggett (W3C/HP)<br>
Ramesh Sarukkai (L&amp;H)<br>
Frank Scahill (BT Labs)<br>
Volker Steinbiss (Philips)<br>
George White (General Magic)</blockquote>

</body>
</html>