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.
2102 lines
90 KiB
2102 lines
90 KiB
<?xml version='1.0' encoding='utf-8' standalone="yes"?>
|
|
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
|
|
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
|
|
|
|
<html xml:lang="en-US" xmlns="http://www.w3.org/1999/xhtml" lang="en-US">
|
|
<head>
|
|
<title>Image Annotation on the Semantic Web</title>
|
|
<meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
|
|
<style type="text/css">
|
|
|
|
.new { color: #FF0000 }
|
|
|
|
.example {font-family: monospace; }
|
|
|
|
.figure {
|
|
font-weight: bold;
|
|
text-align: center; }
|
|
|
|
div.example {
|
|
padding: 1em;
|
|
margin: 0.1em 3.5em 0.1em 0.1em;
|
|
background-color: #efeff5;
|
|
border: 1px solid #cfcfcf; }
|
|
|
|
div.exampleOuter {
|
|
margin: 0em;
|
|
padding: 0em;
|
|
}
|
|
|
|
div.exampleInner {
|
|
color: black;
|
|
background-color: #efeff5;
|
|
border-top-style: double;
|
|
border-top-color: #d3d3d3;
|
|
border-bottom-width: 1px;
|
|
border-bottom-style: double;
|
|
border-bottom-color: #d3d3d3;
|
|
padding: 4px;
|
|
margin: 0em;
|
|
}
|
|
|
|
div.exampleInner pre {
|
|
margin-left: 0em;
|
|
margin-top: 0em;
|
|
margin-bottom: 0em;
|
|
font-family: monospace;
|
|
}
|
|
|
|
div.c1 {text-align:center}
|
|
|
|
</style>
|
|
<link
|
|
href="http://www.w3.org/StyleSheets/TR/W3C-WD.css"
|
|
type="text/css" rel="stylesheet"/>
|
|
</head>
|
|
|
|
<body>
|
|
<div class="head">
|
|
<p>
|
|
<a href="http://www.w3.org/">
|
|
<img alt="W3C" src="http://www.w3.org/Icons/w3c_home"
|
|
height="48" width="72"/>
|
|
</a>
|
|
</p>
|
|
|
|
<h1>Image Annotation on the Semantic Web</h1>
|
|
<h2>W3C Working Draft 22 March 2006</h2>
|
|
|
|
<dl>
|
|
<dt>This version: </dt>
|
|
<dd>
|
|
<a href="http://www.w3.org/TR/2006/WD-swbp-image-annotation-20060322/"
|
|
>http://www.w3.org/TR/2006/WD-swbp-image-annotation-20060322/</a></dd>
|
|
<dt>Latest version: </dt>
|
|
<dd>
|
|
<a href="http://www.w3.org/TR/swbp-image-annotation/"
|
|
>http://www.w3.org/TR/swbp-image-annotation/</a></dd>
|
|
|
|
|
|
<dt>Editors: </dt>
|
|
<dd><a href="http://homepages.cwi.nl/~jrvosse/">Jacco van Ossenbruggen</a>, Center for Mathematics and Computer Science (CWI Amsterdam)</dd>
|
|
<dd><a href="http://homepages.cwi.nl/~troncy/">Raphaël Troncy</a>, Center for Mathematics and Computer Science (CWI Amsterdam)</dd>
|
|
<dd><a href="http://www.image.ntua.gr/~gstam/">Giorgos Stamou</a>, IVML, National Technical University of Athens</dd>
|
|
<dd><a href="http://dl-web.man.ac.uk/~panz/">Jeff Z. Pan</a>, University of Aberdeen (Formerly University of Manchester)</dd>
|
|
|
|
<dt>Contributors: </dt>
|
|
<dd><a href="http://www.mindswap.org/~chris/">Christian Halaschek-Wiener</a>, University of Maryland</dd>
|
|
<dd><a href="mailto:nsimou@image.ece.ntua.gr">Nikolaos Simou</a>, IVML, National Technical University of Athens</dd>
|
|
<dd><a href="mailto:tzouvaras@image.ntua.gr">Vassilis Tzouvaras</a>, IVML, National Technical University of Athens</dd>
|
|
|
|
<dt>  </dt>
|
|
<dd>Also see <a href="#acknowledgments">Acknowledgements</a>.</dd>
|
|
</dl>
|
|
|
|
<p class="copyright"><a href="http://www.w3.org/Consortium/Legal/ipr-notice#Copyright">Copyright</a> © 2006 <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>
|
|
|
|
<hr/>
|
|
</div>
|
|
|
|
<h2>
|
|
<a id="abstract" name="abstract">
|
|
Abstract
|
|
</a>
|
|
</h2>
|
|
|
|
<p>
|
|
Many applications that involve multimedia content make use of
|
|
some form of metadata that describe this content. The goals of
|
|
this document are (i) to explain what the advantages are of
|
|
using Semantic Web languages and technologies for the creation,
|
|
storage, manipulation, interchange and processing of image
|
|
metadata, and (ii) to provide guidelines for doing so. The document
|
|
gives a number of use cases that illustrate ways to exploit
|
|
Semantic Web technologies for image annotation, an overview of RDF
|
|
and OWL vocabularies developed for this task and an overview of
|
|
relevant tools.
|
|
</p>
|
|
|
|
<h2>
|
|
<a id="status" name="status">Status of this document</a>
|
|
</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 document is a First Public Working Draft
|
|
produced by the <a
|
|
href="http://www.w3.org/2001/sw/BestPractices/MM/">Multimedia
|
|
Annotation in the Semantic Web Task Force</a> of the <a
|
|
href="http://www.w3.org/2001/sw/BestPractices/">W3C Semantic
|
|
Web Best Practices & Deployment Working Group</a>. This group
|
|
is part of the <a href="http://www.w3.org/2001/sw/">W3C
|
|
Semantic Web</a> Activity.</p>
|
|
|
|
<p>Discussion of this document
|
|
is invited on the public mailing list <a
|
|
href="mailto:public-swbp-wg@w3.org">public-swbp-wg@w3.org</a>
|
|
(<a
|
|
href="http://lists.w3.org/Archives/Public/public-swbp-wg/">public
|
|
archives</a>). Please start the subject line of the message with
|
|
the text "comments: [MM]".</p>
|
|
|
|
<p>After reviewing comments and further feedback, the Working Group
|
|
may publish new versions of this document or may advance the
|
|
document to Working Group Note.</p>
|
|
|
|
<p> This document was produced by a group operating under the <a href="http://www.w3.org/Consortium/Patent-Policy-20040205/">5 February 2004 W3C Patent Policy</a>. The group does not expect this document to become a W3C Recommendation. This document is informative only. W3C maintains a <a rel="disclosure" href="http://www.w3.org/2004/01/pp-impl/35495/status">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">Essential Claim(s)</a> must disclose the information in accordance with <a href="http://www.w3.org/Consortium/Patent-Policy-20040205/#sec-Disclosure">section 6 of the W3C Patent Policy</a>. </p>
|
|
|
|
<p>Publication as a
|
|
Working Draft does not imply endorsement by the W3C
|
|
Membership. This is a draft document and may be updated,
|
|
replaced or obsoleted by other documents at any time. It is
|
|
inappropriate to cite this document as other than work in
|
|
progress.
|
|
</p>
|
|
|
|
<hr />
|
|
|
|
<h2>
|
|
<a id="roadmap" name="roadmap">
|
|
Document Roadmap
|
|
</a>
|
|
</h2>
|
|
<p>
|
|
After reading this document, readers may
|
|
turn to separate documents discussing individual image
|
|
annotation <a
|
|
href="http://www.w3.org/2001/sw/BestPractices/MM/resources/Vocabularies.html">vocabularies</a>,
|
|
<a
|
|
href="http://www.w3.org/2001/sw/BestPractices/MM/resources/Tools.html">tools</a>,
|
|
and other <a
|
|
href="http://www.w3.org/2001/sw/BestPractices/MM/resources/Resources.html">relevant
|
|
resources</a>.
|
|
Note: many current approaches to image annotation are not based on Semantic
|
|
Web languages. Interoperability between these technologies and
|
|
RDF and OWL-based approaches is <em>not</em> the topic of this document.
|
|
</p>
|
|
|
|
<h2>
|
|
<a id="targetaudience" name="targetaudience">
|
|
Target Audience
|
|
</a>
|
|
</h2>
|
|
|
|
<p>
|
|
This document is target at everybody with an interest in image
|
|
annotation, ranging from non-professional end-users that are
|
|
annotating their personal digital photos to professionals
|
|
working with digital pictures in image and video banks,
|
|
audiovisual archives, museums, libraries, media production and
|
|
broadcast industry, etc.
|
|
</p>
|
|
|
|
<h2>
|
|
<a id="objectives" name="objectives">
|
|
Objectives
|
|
</a>
|
|
</h2>
|
|
|
|
<ul>
|
|
<li>
|
|
To illustrate the benefits of using semantic technologies in image annotations.</li>
|
|
|
|
<li>
|
|
To provide guidelines for applying semantic technologies in this area.</li>
|
|
|
|
<li>
|
|
To collect currently used vocabularies for Semantic Web-based
|
|
image annotations.</li>
|
|
|
|
<li>
|
|
To provide use cases with examples of Semantic Web-based
|
|
annotations.</li>
|
|
</ul>
|
|
|
|
<div class="toc">
|
|
<h2 class="notoc">
|
|
<a id="contents" name="contents">Table of Contents</a>
|
|
</h2>
|
|
|
|
<ul id="toc" class="toc">
|
|
<li class="tocline"><a href="#introduction"><b>1. Introduction</b></a>
|
|
<ul class="toc">
|
|
<li class="tocline"><a href="#annot_intro">1.1 Image Annotation Issues</a></li>
|
|
<li class="tocline"><a href="#semweb_intro">1.2 Semantic Web Basics</a></li>
|
|
</ul>
|
|
</li>
|
|
<li class="tocline"><a href="#use_cases"><b>2. Use Cases</b></a>
|
|
<ul class="toc">
|
|
<li class="tocline"><a href="#world_images">2.1. World Images</a></li>
|
|
<li class="tocline"><a href="#culture_images">2.2. Culture Images</a></li>
|
|
<li class="tocline"><a href="#media_images">2.3. Media</a></li>
|
|
<li class="tocline"><a href="#scientific_images">2.4. Scientific Images</a></li>
|
|
</ul>
|
|
</li>
|
|
<li class="tocline"><a href="#vocabularies"><b>3. Vocabularies for Image Annotation</b></a></li>
|
|
<li class="tocline"><a href="#tools"><b>4. Available Tools for Semantic Image Annotation</b></a></li>
|
|
<li class="tocline"><a href="#examples"><b>5. Example Solutions to the Use Cases</b></a>
|
|
<ul class="toc">
|
|
<li class="tocline"><a href="#solution_personal">5.1. Use Case: Management of Personal Digital Photo Collections</a></li>
|
|
<li class="tocline"><a href="#solution_culture">5.2. Use Case: Cultural Heritage</a></li>
|
|
<li class="tocline"><a href="#solution_TVarchive">5.3. Use Case: Television News Archive</a></li>
|
|
<li class="tocline"><a href="#solution_NASA">5.4. Use Case: large-scale image collections at NASA</a></li>
|
|
</ul>
|
|
</li>
|
|
<li class="tocline"><a href="#conclusions"><b>6. Conclusions</b></a></li>
|
|
<li class="tocline"><a href="#references"><b>References</b></a></li>
|
|
<li class="tocline"><a href="#acknowledgments"><b>Acknowledgments</b></a></li>
|
|
</ul>
|
|
</div>
|
|
|
|
<h2>
|
|
<a name="introduction">
|
|
1. Introduction
|
|
</a>
|
|
</h2>
|
|
|
|
<p>
|
|
The need for annotating digital image data is recognized in a
|
|
wide variety of different applications, covering both
|
|
professional and personal usage of image data. At the time of writing,
|
|
most work done in this area does not use semantic-based technologies partly
|
|
because of the differences between the multimedia and the web communities
|
|
and their underlying standardization organizations.
|
|
This document explains the advantages of
|
|
using Semantic Web languages and technologies for image
|
|
annotations and provides guidelines for doing so. It is
|
|
organized around a number of representative use cases, and a
|
|
description of Semantic Web vocabularies and tools that could be
|
|
used to help accomplish the task mentioned in the uses cases.
|
|
The remainder of this introductory section first gives an
|
|
overview of image annotation in general, followed by a short
|
|
description of the key Semantic Web concepts that are relevant
|
|
for image annotation.
|
|
</p>
|
|
|
|
<h3>
|
|
<a name="annot_intro">
|
|
1.1 Image Annotation Issues
|
|
</a>
|
|
</h3>
|
|
|
|
<p>
|
|
Annotating images on a small scale for personal usage can be
|
|
relatively simple. The reader should be warned, however, that
|
|
large scale, industrial strength image annotation is notoriously
|
|
complex. Trade offs along several dimensions make the professional
|
|
multimedia annotations difficult:
|
|
</p>
|
|
|
|
<ol>
|
|
<li>
|
|
<p>
|
|
<em>Production versus post-production annotation</em>
|
|
</p>
|
|
<p>
|
|
A general rule is that it is much easier to annotate earlier rather than later.
|
|
Typically, most of the information that is needed for
|
|
making the annotations is available during production
|
|
time. Examples include time and date, lens settings and
|
|
other EXIF metadata added to JPEG images by most digital
|
|
cameras at the time a picture is taken, experimental data in
|
|
scientific and medical images, information from scripts,
|
|
story boards and edit decision lists in creative industry,
|
|
etc. Indeed, maybe the single most best practice in image
|
|
annotation is that in general, adding metadata during the
|
|
production process is much cheaper and yields higher quality
|
|
annotations than adding metadata in a later stage (such as
|
|
by automatic analysis of the digital artifact or by manual
|
|
post-production data).
|
|
</p>
|
|
</li>
|
|
<li>
|
|
<p>
|
|
<em>
|
|
Generic vs task-specific annotation
|
|
</em>
|
|
</p>
|
|
<p>
|
|
Annotating images without having a specific goal or task in
|
|
mind is often not cost effective: after the target
|
|
application has been developed, it turns out that images
|
|
have been annotated using the wrong type of information, or
|
|
on the wrong abstraction level, etc. Redoing the annotations
|
|
is then an unavoidable, but costly solution. On the other
|
|
hand, annotating with <em>only</em> the target application
|
|
in mind may also not be cost effective. The annotations may
|
|
work well with that one application, but if the same
|
|
metadata is to be reused in the context of other
|
|
applications, it may turn out to be too specific, and
|
|
unsuited for reuse in a different context. In most
|
|
situations the range of applications in which the metadata
|
|
will be used in the future is unknown at the time of
|
|
annotation. When lacking a crystal ball, the best the
|
|
annotator can do in practice is use an approach that is
|
|
sufficiently specific for the application under
|
|
development, while avoiding unnecessary application-specific
|
|
assumptions as much as possible.
|
|
</p>
|
|
</li>
|
|
<li>
|
|
<p>
|
|
<em>
|
|
Manual versus automatic annotation and the "Semantic Gap"
|
|
</em>
|
|
</p>
|
|
<p>
|
|
In general, manual annotation can provide image descriptions
|
|
at the right level of abstraction. It is, however, time
|
|
consuming and thus expensive. In addition, it proves to be
|
|
highly subjective: different human annotators tend to "see"
|
|
different things in the same image. On the other hand,
|
|
annotation based on automatic feature extraction is
|
|
relatively fast and cheap, and can be more systematic. It tends
|
|
to result, however, in image descriptions that are too low
|
|
level for many applications. The difference between the low
|
|
level feature descriptions provided by image analysis tools
|
|
and the high level content descriptions required by the
|
|
applications is often referred to, in the literature, as the
|
|
<em>Semantic Gap</em>. In the remainder, we will discuss use
|
|
cases, vocabularies and tools for both manual and automatic
|
|
image annotation.
|
|
</p>
|
|
</li>
|
|
|
|
<li>
|
|
<p>
|
|
<em>
|
|
Different types of metadata
|
|
</em>
|
|
</p>
|
|
<p>
|
|
While various classifications of metadata have been described in
|
|
the literature, every annotator should at least be aware of the
|
|
difference between annotations describing properties of the
|
|
image itself, and those describing the subject matter of the
|
|
image, that is, the properties of the objects, persons or
|
|
concepts depicted by the image. In the first category, typical
|
|
annotations provide information about title, creator,
|
|
resolution, image format, image size, copyright, year of
|
|
publication, etc. Many applications use a common, predefined
|
|
and relatively small vocabulary defining such properties.
|
|
Examples include the <a href="#DublinCore">Dublin Core</a> and
|
|
<a href="#VraCore">VRA Core</a> vocabularies. The second
|
|
category describes what is depicted by the image, which can vary
|
|
wildly with the type of image at hand. In many applications, it
|
|
is also useful to distinguish between objective observations
|
|
('the person in the white shirt moves his arm from left to
|
|
right') versus subjective interpretations ('the person seems to
|
|
perform a martial arts exercise). As a result, one sees a large
|
|
variation in vocabularies used for this purpose. Typical
|
|
examples vary from domain-specific vocabularies (for example,
|
|
with terms that are very specific for astronomy images, or sport
|
|
images, etc) to domain-independent ones (for example, a
|
|
vocabulary with terms that are sufficiently generic to describe
|
|
any news photo). In addition, vocabularies tend to differ in
|
|
size, granularity, formality etc. In the remainder, we discuss
|
|
the above metadata categories. Note that in the first type it
|
|
is not uncommon that a vocabulary only defines the properties
|
|
and defers the definitions of the values of those properties to
|
|
another vocabulary. This is true, for example, for both Dublin
|
|
Core and VRA Core. This means that, typically, in order to
|
|
annotate a single image one needs terms from multiple
|
|
vocabularies.
|
|
</p>
|
|
</li>
|
|
|
|
<li>
|
|
<p>
|
|
<em>
|
|
Lack of Syntactic and Semantic Interoperability
|
|
</em>
|
|
</p>
|
|
<p>
|
|
Many different file formats and tools for image annotations
|
|
are currently in use. Reusing metadata developed for one set
|
|
of tools in another tool is often hindered by a lack of
|
|
interoperability. First, different tools use different file
|
|
formats, so tool A may not be able to read in the metadata
|
|
provided by tool B (syntax-level interoperability). Solving
|
|
the problem is relatively easy if the inner structure of
|
|
both file formats are known by developing a conversion tool.
|
|
Second, tool A may assign a different meaning to the same
|
|
annotation as tool B does (semantic
|
|
interoperability). Solving this problem is much harder, and a first
|
|
step to provide a solution is to require that the
|
|
vocabulary used be explicitly defined for both tools.
|
|
</p>
|
|
</li>
|
|
</ol>
|
|
|
|
<h3>
|
|
<a name="semweb_intro">
|
|
1.2 Semantic Web Basics
|
|
</a>
|
|
</h3>
|
|
|
|
<p>
|
|
This section briefly describe the role of Semantic Web technologies in image annotations. The aim of the Semantic Web is to augment the
|
|
existing Web so that resources (Web pages, images etc.) are more easily interpreted by programs (or "intelligent agents"). The idea is to
|
|
associate Web resources with semantic categories which describe the contents and/or functionalities of Web resources.
|
|
</p>
|
|
|
|
<p>
|
|
Annotations alone do not establish the semantics of what is being marked-up. One way generally followed
|
|
to introduce semantics to annotations is to get an agreement to carefully define what a set of
|
|
concepts are and what terms have to be used for them.
|
|
</p>
|
|
|
|
<p>
|
|
This agreement can be only "informal", that is, relies on natural language for defining the meaning
|
|
of a set of information properties. For example, the <a href="#DublinCore">Dublin Core</a> Metadata
|
|
Element Set provides 15 "core" information properties, such as "Title", "Creator", "Date", with
|
|
descriptive semantic definitions (in natural language). One can use these information properties in,
|
|
e.g., <a href="#RDF">RDF</a> or META tags of HTML.
|
|
</p>
|
|
|
|
<p>
|
|
For example, the following RDF/XML code represents the statements "there is an image <tt>Ganesh.jpg</tt> created by <tt>Jeff Z. Pan</tt> and whose
|
|
title is <tt>An image about the Elephant Ganesh</tt>. The first four lines define the <a href="#XML-NS">XML namespaces</a> used in this
|
|
description. A good starting point for having more information on RDF is the <a href="#RDF-Primer">RDF Primer</a>.
|
|
</p>
|
|
|
|
<table width="90%" border="0" cellspacing="20" bgcolor="#EEEEEE">
|
|
<tr>
|
|
<td>
|
|
<pre>
|
|
<rdf:RDF xml:base="http://example.org/"
|
|
xmlns="http://example.org/"
|
|
xmlns:dc="http://purl.org/dc/elements/1.1/"
|
|
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
|
|
|
|
<rdf:Description rdf:about="Ganesh.jpg"/>
|
|
<dc:title>An image about the Elephant Ganesh</dc:title>
|
|
<dc:creator>Jeff Z. Pan</dc:creator>
|
|
</rdf:Description>
|
|
</rdf:RDF></pre>
|
|
</td>
|
|
</tr>
|
|
</table>
|
|
|
|
<p>
|
|
A complementary approach is to also use ontologies to specify formally the meaning of Web resources
|
|
and thus get a "formal" agreement. <em>Ontology</em> is a term borrowed from philosophy
|
|
that refers to the science of describing the kinds of entities in the world and how they are related.
|
|
In computer science, ontology is, in general, a representation of a shared conceptualization of a
|
|
specific domain. It provides a shared and common <em>vocabulary</em>, including important concepts,
|
|
properties, their definitions and <em>constraints</em>, sometimes referred to as background assumptions
|
|
regarding the intended meaning of the vocabulary, used in a domain that can be communicated between
|
|
people and heterogeneous, distributed application systems. The (formal) ontology approach, though more
|
|
difficult to develop, is more powerful than the informal-only agreement approach because users can
|
|
thoroughly define the vocabulary using axioms expressed in a logic language and machine can use this
|
|
formal meaning for reasoning, completing and validating the annotations. Ideally, the concepts and
|
|
properties of an ontology should have both formal definitions and natural language descriptions to be
|
|
unambiguously used by humans and software applications.
|
|
</p>
|
|
|
|
<p>
|
|
There exists a standard Semantic Web Ontology Language <a href="#OWL">OWL</a>, which is a W3C
|
|
recommendation. We provide below an example of this language in its RDF/XML syntax.
|
|
Given that there exists a <tt>Image</tt> class and a <tt>hasSize</tt> property in an ontology, one can
|
|
use the following OWL statements to define a new OWL class called <tt>BigImage</tt> as the set of all
|
|
members of the class <tt>Image</tt> such that the size of the image is equal to <tt>Big</tt>.
|
|
For more information, the <a href="#OWL-Guide">OWL Guide</a> provides a good overview of the OWL
|
|
language.
|
|
</p>
|
|
|
|
<table width="90%" border="0" cellspacing="20" bgcolor="#EEEEEE">
|
|
<tr>
|
|
<td>
|
|
<pre>
|
|
<rdf:RDF xml:base="http://example.org/"
|
|
xmlns="http://example.org/"
|
|
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
|
|
xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema##">
|
|
xmlns:owl="http://www.w3.org/2002/07/owl#"
|
|
|
|
<owl:Class rdf:about="BigImage"/>
|
|
<owl:intersectionOf rdf:parseType="Collection">
|
|
<owl:Class rdf:about="#Image">
|
|
<owl:Restriction>
|
|
<owl:onProperty rdf:resource="#hasSize">
|
|
<owl:cardinality rdf:datatype="&xsd;nonNegativeInteger">1</owl:cardinality>
|
|
<owl:allValueFrom rdf:resource="#Big">
|
|
</owl:Restriction>
|
|
</owl:intersectionOf>
|
|
</owl:Class>
|
|
</rdf:RDF></pre>
|
|
</td>
|
|
</tr>
|
|
</table>
|
|
|
|
<p>
|
|
The next section presents some representative use cases that highlight some requirements for image
|
|
annotation tools, vocabularies, and practices.
|
|
</p>
|
|
|
|
<h2>
|
|
<a name="use_cases">
|
|
2. Use Cases
|
|
</a>
|
|
</h2>
|
|
<p>
|
|
Image annotation is relevant in a wide range of domains,
|
|
organizations and applications; it cannot be covered in a
|
|
single document such as this. This document, instead, describes a number of use cases
|
|
that are intended as a representative set of
|
|
examples. These use cases will be used later to discuss the
|
|
vocabularies and tools that are relevant for image annotation on
|
|
the Semantic Web. Example scenarios are given in <a
|
|
href="#examples">Section 5</a>.
|
|
</p>
|
|
|
|
<p>
|
|
The use cases are organized in four categories, which reflect
|
|
either the topics depicted by the images or their usage community.
|
|
These criteria often determine the tools and vocabularies used in the annotation process.
|
|
</p>
|
|
|
|
<h3>
|
|
<a name="world_images">
|
|
2.1 World Images
|
|
</a>
|
|
</h3>
|
|
<p>
|
|
This section provides two use cases with images that could
|
|
potentially depict any subject: management of a personal photo
|
|
collection and that of a news press photo bank. The other use
|
|
cases will focus on images from a specific domain.
|
|
</p>
|
|
|
|
<h4>
|
|
<a name="photo_collection">
|
|
Use case: Management of Personal Digital Photo Collections
|
|
</a>
|
|
</h4>
|
|
|
|
<p>
|
|
Many personal users have thousands of digital photos from vacations, parties, traveling,
|
|
conferences, everyday life etc. Typically, the photos are stored on personal computer hard drives
|
|
in a simple directory structure without any metadata. The user wants generally to easily access
|
|
this content, view it, use it in his homepage, create presentations, make part of it
|
|
accessible for other people or even sell part of it to image banks. Too often, however, the only
|
|
way for this content to be accessed is by browsing the directories, their name providing usually
|
|
the date and the description with one or two words of the original event captured by
|
|
the specific photos. Obviously, this access becomes more and more difficult as the number of
|
|
photos increases and the content becomes quickly unused in practice.
|
|
More sophisticated users leverage simple photo organizing tools allowing them to provide
|
|
keyword metadata, possibly along with a simple taxonomy of categories. This is a first step
|
|
towards a semantically-enabled solution. <a href="#solution_personal">Section
|
|
5.1</a> provides an example scenario for this use case using Semantic Web technologies.
|
|
</p>
|
|
|
|
<h4>
|
|
<a name="photo_bank">
|
|
Use case: Press Photo Bank
|
|
</a>
|
|
</h4>
|
|
<p>
|
|
<!--
|
|
TO DO: IPTC / News / Sport / Entertainment. e.g Corbis, Associated Press, Reuters
|
|
-->
|
|
</p>
|
|
|
|
<h3>
|
|
<a name="culture_images">
|
|
2.2 Culture Images
|
|
</a>
|
|
</h3>
|
|
<p>
|
|
This section contains a single use case from the cultural
|
|
heritage domain. This domain is characterized by a long
|
|
tradition in describing images, with many standardized methods
|
|
and vocabularies.
|
|
</p>
|
|
|
|
<h4>
|
|
<a name="cultural_heritage">
|
|
Use case: Cultural Heritage
|
|
</a>
|
|
</h4>
|
|
<p>
|
|
Let us imagine that a museum in fine arts has asked a specialized company to produce
|
|
high resolution digital scans of the most important art works of
|
|
their collections. The museum's quality assurance requires the
|
|
possibility to track when, where and by whom every scan was
|
|
made, with what equipment, etc. The museum's internal IT
|
|
department, maintaining the underlying image database, needs the
|
|
size, resolution and format of every resulting image. It also
|
|
needs to know the repository ID of the original work of art. The
|
|
company developing the museum's website additionally requires
|
|
copyright information (that varies for every scan, depending on
|
|
the age of the original work of art and the collection it
|
|
originates from). It also want to give the users of the website
|
|
access to the collection, not only based on the titles of the
|
|
paintings and names of their painters, but also based on the
|
|
topics depicted ('sun sets'), genre ('self portraits'), style
|
|
('post-impressionism'), period ('fin de siècle'), region ('west
|
|
European'). <a href="#solution_culture">Section 5.2</a>
|
|
shows how all these requirements can be fulfilled using Semantic Web technologies.
|
|
</p>
|
|
|
|
<h3>
|
|
<a name="media_images">
|
|
2.3 Media
|
|
</a>
|
|
</h3>
|
|
<p>
|
|
The use case developed in this section is mainly targeted at media
|
|
professionals, and less to the general public. Typical requests
|
|
are characterized by very detailed queries, not only about the
|
|
content of images, but also about the media specific details such as
|
|
camera angle, lens settings etc.
|
|
</p>
|
|
|
|
<h4>
|
|
<a name="television_archive">
|
|
Use case: Television Archive
|
|
</a>
|
|
</h4>
|
|
|
|
<p>
|
|
Audiovisual archive centers are used to manage very large
|
|
multimedia databases. For instance, INA, the French Audiovisual
|
|
National Institute, has been archiving TV documents for 50 years
|
|
and radio documents for 65 years and stores more than 1 million
|
|
hours of broadcast programs. The images and sound archives kept
|
|
at INA are either intended for professional use (journalists,
|
|
film directors, producers, audiovisual and multimedia
|
|
programmers and publishers, in France and worldwide) or
|
|
communicated for research purposes (for a public of students,
|
|
research workers, teachers and writers). In order to allow an
|
|
efficient access to the data stored, most of the parts of these
|
|
video documents are described and indexed by their content. The
|
|
global multimedia information system should then be fine-grain
|
|
enough detailed to support some very complex and precise
|
|
queries. For example, a journalist or a film director client
|
|
might ask for an excerpt of a previously broadcasted program
|
|
showing the first goal of a given football player in its
|
|
national team, scored with its head. The query could
|
|
additionally contain some more technical requirements such that
|
|
the goal action should be available according to both the front
|
|
camera view and the reverse angle camera view. Finally, the
|
|
client might or might not remember some general information
|
|
about this football game, such that the date, the place and the
|
|
final score. <a href="#solution_TVarchive">Section 5.3</a>
|
|
gives a possible solution for this use case using Semantic Web technologies.
|
|
</p>
|
|
|
|
<h3>
|
|
<a name="scientific_images">
|
|
2.4 Scientific Images
|
|
</a>
|
|
</h3>
|
|
<p>
|
|
This section presents two use cases from the scientific domain.
|
|
Typically here, images are annotated using large and complex
|
|
ontologies.
|
|
</p>
|
|
|
|
<h4>
|
|
<a name="large_collection">
|
|
Use Case: Large-scale Image Collections at NASA
|
|
</a>
|
|
</h4>
|
|
|
|
<p>
|
|
Many organizations maintain extremely large-scale image
|
|
collections. The National Aeronautics and Space Administration
|
|
(NASA), for example, has hundreds of thousands of
|
|
images, stored in different formats, levels of availability and
|
|
resolution, and with associated descriptive information at
|
|
various levels of detail and formality. Such an organization
|
|
also generates thousands of images on an ongoing basis that are
|
|
collected and cataloged. Thus, a mechanism is needed to catalog
|
|
all the different types of image content across various
|
|
domains. Information about both the image itself (e.g., its
|
|
creation date, dpi, source) and about the specific content of
|
|
the image is required. Additionally, the associated metadata
|
|
must be maintainable and extensible so that associated
|
|
relationships between images and data can evolve
|
|
cumulatively. Lastly, management functionality should provide
|
|
mechanisms flexible enough to enforce restriction based on
|
|
content type, ownership, authorization, etc.
|
|
<a href="#solution_NASA">Section 5.4</a>
|
|
gives an example solution for this use case.
|
|
</p>
|
|
|
|
<h4>
|
|
<a name="medical_images">
|
|
Use Case: Bio-Medical Images
|
|
</a>
|
|
</h4>
|
|
|
|
<p>
|
|
</p>
|
|
|
|
<h2>
|
|
<a name="vocabularies">
|
|
3. Vocabularies for Image Annotation
|
|
</a>
|
|
</h2>
|
|
|
|
<p>
|
|
Choosing which vocabularies to use for annotating image is a key
|
|
decision in an annotation project. Typically, one needs more
|
|
than a single vocabulary to cover the different relevant aspects
|
|
of the images. A separate document named
|
|
<a href="http://www.w3.org/2001/sw/BestPractices/MM/resources/Vocabularies.html">Vocabularies
|
|
Overview</a> discusses a number of individual vocabularies that
|
|
are relevant for images annotation. The remainder of this
|
|
section discusses more general issues.
|
|
</p>
|
|
|
|
<p>
|
|
Many of the relevant vocabularies have been developed prior to
|
|
the Semantic Web, and <a
|
|
href="http://www.w3.org/2001/sw/BestPractices/MM/resources/Vocabularies.html">Vocabularies
|
|
Overview</a> lists
|
|
many translations of such vocabularies to RDF or OWL. Most
|
|
notably, the key International Standard in this area, the <a
|
|
href="#MPEG-7">Multimedia Content Description</a> standard,
|
|
widely known as MPEG-7, is defined using XML Schema. At the
|
|
time of writing, there is no commonly accepted mapping from the
|
|
XML Schema definitions in the standard to RDF or OWL. Several
|
|
alternative mappings, however, have been developed so far and
|
|
are discussed in the overview.
|
|
</p>
|
|
|
|
<p>
|
|
Another relevant vocabulary is the <a href="#VraCore">VRA
|
|
Core</a>. Where the <a href="#DublinCore">Dublin Core</a> (DC)
|
|
specifies a small and commonly used vocabulary for on-line
|
|
resources in general, VRA Core defines a similar set targeted
|
|
especially at visual resources, specializing the DC elements.
|
|
Dublin Core and VRA Core both refer to terms in their
|
|
vocabularies as <em>elements</em>, and both use
|
|
<em>qualifiers</em> to refine elements in similar way. All the
|
|
elements of VRA Core have either direct mappings to comparable
|
|
fields in Dublin Core or are defined as specializations of one
|
|
or more DC elements. Furthermore, both vocabularies are defined
|
|
in a way that abstracts from implementation issues and
|
|
underlying serialization languages. A key difference, however,
|
|
is that for Dublin Core, there exists a commonly accepted
|
|
mapping to RDF, along with the associated schema. At the time of
|
|
writing, this is not the case for VRA Core, and the overview
|
|
discusses the pros and cons of the alternative mappings.
|
|
</p>
|
|
|
|
<p>
|
|
Many annotations on the Semantic Web are about an entire
|
|
resource. For example, a <tt><dc:title></tt> property
|
|
applies to the entire document. For images and other multimedia
|
|
documents, one often needs to annotate a specific part of a
|
|
resource (for example, a region in an image). Sharing the
|
|
metadata dealing with the localization of some specific part of
|
|
multimedia content is important since it allows to have multiple
|
|
annotations (potentially from multiple users) referring to the
|
|
same content.
|
|
</p>
|
|
|
|
<!--
|
|
<p>
|
|
[TO DO: Discuss and give examples of two possible solutions:]
|
|
</p>
|
|
-->
|
|
|
|
<ol>
|
|
<li>
|
|
Ideally, the target image already specifies this specific
|
|
part, using a name that is addressable in the URI fragment
|
|
identifier (this can be done, for example, in SVG).
|
|
</li>
|
|
<li>
|
|
Otherwise the region needs to be described in the metadata itself, as
|
|
it is done in MPEG-7.
|
|
</li>
|
|
</ol>
|
|
|
|
<!--
|
|
<p>
|
|
[TO DO: Add concluding remarks / too abrupt currently]
|
|
</p>
|
|
-->
|
|
<h2>
|
|
<a name="tools">
|
|
4. Available Tools for Semantic Image Annotation
|
|
</a>
|
|
</h2>
|
|
|
|
<p>
|
|
Among the numerous tools used for image archiving and description, some of them
|
|
may be used for semantic annotation. The aim of this section is to
|
|
identify some key characteristics of semantic image annotation tools, so as to provide
|
|
some guidelines for their proper use. Using these characteristics as criteria, users of these
|
|
tools could choose the most appropriate for a specific application.
|
|
</p>
|
|
|
|
<!--
|
|
<p>
|
|
TO DO: clarify here that we would like to discuss what are the abilities of the tools:
|
|
can they handle different type of content ? do they allow fine-grained annotations ? etc ...
|
|
Obviously, some of these characteritics are intrinsically related to the images themselves,
|
|
or to what users need to do with them, but at the end, we should emphasize that the main
|
|
bottleneck will be what the tools can actually do (not much currently :-( !
|
|
|
|
MIKE: Fair point. It would be good to say something about this to avoid confusion.
|
|
</p>
|
|
-->
|
|
|
|
<p>
|
|
<strong>Type of Content.</strong> A tool can annotate different type of content.
|
|
Usually, the raw content is an image, whose format can be jpg, png, tif, etc. but there
|
|
are also tools that can annotate videos as well.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>Type of Metadata.</strong> An annotation can be targeted for different use.
|
|
Following the categorization
|
|
provided by <a href="http://sunsite.berkeley.edu/moa2/wp-v2.html">The Making of
|
|
America II project</a>, the metadata can be <em>descriptive</em> (for description
|
|
and identification of information), <em>structural</em>
|
|
(for navigation and presentation), or <em>administrative</em> (for management and
|
|
processing). Most of the tools can be used in order to provide
|
|
descriptive metadata and for some of them, the user can also provide structural and
|
|
administrative information.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>Format of Metadata.</strong> An annotation can be expressed in different format.
|
|
This format is important since it should ensure
|
|
interoperability with other (semantic web) applications. MPEG-7 is often used as
|
|
the metadata format for exchanging automatic analysis results whereas OWL and RDF are
|
|
better appropriate in the Semantic Web world.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>Annotation level.</strong> Some tools give to the user the
|
|
opportunity to annotate an image using vocabularies while others allow free text
|
|
annotation only. When ontologies are used (in RDF or OWL format),
|
|
the annotation level is considered to be controlled since the semantics is generally
|
|
provided in a more formal way, whereas if they are not, the annotation level is
|
|
considered to be free.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>Client-side Requirement.</strong> This characteristic refers to
|
|
whether users can use a Web browser to access the service(s) or need to install
|
|
a stand-alone application.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>License Conditions.</strong> Some of the tools are open source while some
|
|
others are not. It is important for the user and for potential researchers and
|
|
developers in the area of multimedia annotation to know this issue
|
|
before choosing a particular tool.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>Collaborative or individual.</strong> This characteristic refers
|
|
to the possible usage of the tool as an annotation framework for web-shared image
|
|
databases or as an individual user multimedia content annotation tool.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>Granularity.</strong> Granularity specifies whether annotation is
|
|
segment based or file based. This is an important characteristic since in some applications,
|
|
it could be crucial to provide the structure of the image. For example, it is
|
|
useful to provide annotations for different areas of the image, describing several cues of
|
|
information (like a textual part or sub-images) or defining and describing different objects
|
|
visualized in the image (e.g. people).
|
|
</p>
|
|
|
|
<p>
|
|
<strong>Threaded or unthreaded.</strong> This characteristic refers to the
|
|
ability of the tool to respond or add to a previous annotation and to stagger/structure
|
|
the presentation of annotations to reflect this.
|
|
</p>
|
|
|
|
<p>
|
|
<strong>Access control.</strong> This refers to the
|
|
access provided for different users to the metadata. For example, it is important to
|
|
distinguish between users that have simple access (just view) and users that have full
|
|
access (view or change).
|
|
</p>
|
|
|
|
<p>
|
|
Concluding, the appropriateness of a tool depends on the nature of annotation that the user
|
|
requires and cannot be predetermined. A separate web page is maintained with
|
|
<a href="http://www.w3.org/2001/sw/BestPractices/MM/resources/Tools.html">Semantic Web
|
|
Image Annotation Tools</a>, and categorizes most of the annotation tools found in the Internet,
|
|
according to the characteristics described above. Any comments, suggestions or new tools
|
|
annoucements will be added to this separate document. The tools can be used for different
|
|
types of annotations, depending on the use cases, as shown in the following section.
|
|
</p>
|
|
|
|
<!-- EXAMPLES -->
|
|
<h2>
|
|
<a name="examples">
|
|
5. Example Solutions to the Use Cases
|
|
</a>
|
|
</h2>
|
|
<p>
|
|
This section describes possible scenarios for how
|
|
Semantic Web technology could be used for supporting the
|
|
use cases presented in <a href="#use_cases">Section 2</a>.
|
|
These scenarios are provided purely as illustrative examples and do not imply
|
|
endorsement by the W3C membership or the Semantic Web Best
|
|
Practices and Deployment Working Group.
|
|
</p>
|
|
|
|
<h3>
|
|
<a name="solution_personal">
|
|
5.1 Use Case: Management of Personal Digital Photo Collections
|
|
</a>
|
|
</h3>
|
|
|
|
<div style="float: right; width: 300px; border: 1px solid gray; padding: 1%; margin: 1%">
|
|
<a href="Personal.jpg">
|
|
<img style="width: 300px;"
|
|
src="Personal.jpg"
|
|
alt="A photo from a personal collection"/></a>
|
|
<br/>
|
|
A photo from a personal collection
|
|
</div>
|
|
|
|
<h4 id="personal_solution">Possible Semantic Web-based solution</h4>
|
|
<p>
|
|
The solution of the use case described in <a href="#world_images">Section 2.1</a> requires
|
|
the use of multiple
|
|
vocabularies. The potential domain of a photo from a personal
|
|
digital collection is very wide, and may include sports,
|
|
entertainment, sightseeing etc. In order to solve this use case the
|
|
information that a user needs to know about the image has to be
|
|
taken into account for a appropriate selection of vocabularies. The
|
|
use case requires creating semantic labels and associate them with
|
|
the photo. Semantic labels may refer to both media and content
|
|
type annotations. The examples cover three different approaches:
|
|
Manual, Semi-Automatic, and Automatic. Each approach has advantages
|
|
and disadvantages and each one requires different solutions.
|
|
</p>
|
|
|
|
|
|
<h5>Manual Annotation</h5>
|
|
<p>
|
|
Manual annotation potentially offers the most accurate information,
|
|
but it is the most time demanding and thus expensive. In manual
|
|
annotation, there is typically no need for creating comprehensive
|
|
annotations based on media features (e.g. low-level image
|
|
characteristics also known as <em>visual descriptors</em>) since
|
|
most users are not interested in querying the image database using
|
|
low-level features such as shape, texture, color histograms etc.
|
|
However, for most applications, some minimal media type information
|
|
is needed such as the type of the image (i.e. jpeg, tiff etc.) or
|
|
the resolution of the image. In addition, provenance information
|
|
regarding the date created, the creator, the thematic category
|
|
etc. is also common. VRA <a href="#VraRDF">[VRA in RDF/OWL]</a> can
|
|
be used to describe the above information.
|
|
</p>
|
|
<p>
|
|
Regarding the actual content of the image, various vocabularies can
|
|
be used depending on the respective thematic category. The example
|
|
shows a photo that has content from the beach holidays thematic
|
|
category. For this reason, a beach ontology
|
|
and the PhotoStuff image annotator <a href="#PS">[PhotoStuff]</a>
|
|
can be used to describe the image content.
|
|
</p>
|
|
|
|
|
|
<h5>Semi-Automatic Annotation</h5>
|
|
<p>
|
|
Semi-Automatic Annotation assists the manual annotation to extract
|
|
higher-level, semantic labels (or vice versa). Image analysis
|
|
tools such as image segmentation and object recognition tools are
|
|
based on lower level aspects of the media. As a result, a more
|
|
extensive set of lower level media type descriptors is needed in
|
|
this approach. The current trend in the multimedia community is
|
|
that the combination of image analysis tools with
|
|
multimedia-specific and domain-specific vocabularies is shifting the
|
|
image analysis, recognition and retrieval processes to a more
|
|
semantic level.
|
|
</p>
|
|
|
|
<p>
|
|
Using the above holiday beach example, in order to
|
|
semi-automatically annotate the image, low-level image concepts and
|
|
relations are needed (color, shape, texture etc.). The MPEG-7
|
|
visual part <a href="#MPEG-7">[MPEG-7]</a> is an appropriate
|
|
framework for the representation of such features. For this reason,
|
|
a Visual Descriptor Ontology (VDO) <a href="#VDO">[VDO]</a> in
|
|
combination with the beach domain ontology can be used to assign
|
|
visual descriptors to domain concepts in order to be automatically
|
|
recognized and thus annotated. For example, the M-OntoMat
|
|
Annotizer can be used to manually segment objects that have a
|
|
semantic meaning, then extract the respective visual descriptors
|
|
and store them as prototype instances in a predefined domain
|
|
ontology (beach ontology). In addition, reasoning support is also
|
|
required in the semi-automatic process. Using reasoning tools,
|
|
higher level concepts and events can be recognized in the image.
|
|
Multimedia reasoning tools require spatio-temporal knowledge about
|
|
the objects of the image (e.g. a person consists of a body, two
|
|
hands, two legs and a head; or: the sky is over the sea etc.). An
|
|
example of visual descriptors in association with domain concepts
|
|
using M-OntoMat Annotizer is shown in the RDF graph below (<a
|
|
href="#figure1">Figure 1</a>). The RDF code can be found <a
|
|
href="PersonalContent_M_Ontomat.rdf">here</a>.
|
|
</p>
|
|
|
|
<div style="text-align: center">
|
|
<img alt="An RDF Graph Describing the association of MPEG-7 Visual descriptors"
|
|
style="width: 80%;"
|
|
src="PersonalRDFDiagram.png"/>
|
|
<br/>
|
|
<a id="figure1" name="figure1">
|
|
Figure 1: An RDF Graph Describing the association of MPEG-7 Visual
|
|
descriptors with the domain concept "sand"</a>
|
|
</div>
|
|
|
|
|
|
<h5>Automatic Annotation</h5>
|
|
<p>
|
|
Automatic Annotation means that no user involvement is needed, and
|
|
thus is time and cost effective. However, even with perfect image
|
|
segmentation, person detection and object recognition, a tool will
|
|
not recognize events such as "Katerina's' holidays in
|
|
Thailand". In the beach holiday example, more vocabularies are
|
|
needed such as a context ontology for acquiring the context of the
|
|
image (e.g. automatically detect that the image is about holidays
|
|
in beaches and not in mountains) in
|
|
order to automatically annotating the image. Also, automation is
|
|
needed in creating the prototype instances using the VDO, the
|
|
domain ontologies and the M-OntoMat Annotizer in order to automatically
|
|
segment regions that may have semantic meaning and then extract and store
|
|
the visual descriptors. Such an advanced approach is beyond the
|
|
scope of this deliverable.
|
|
</p>
|
|
|
|
|
|
|
|
<h4 id="personalconclusion">Conclusion and discussion</h4>
|
|
<p>
|
|
The example solution shows that even the manual annotation
|
|
is non-trivial. It is difficult to provide a unified way to
|
|
annotate personal photos. The context of the photo indicates which ontology
|
|
must be used in the annotation process. In the above example, a beach
|
|
domain ontology is used since the context of the photo is summer holidays. Apart from domain
|
|
specific ontologies, media type ontologies and a photo annotation tool are required to
|
|
complete the annotation.
|
|
</p>
|
|
<p>
|
|
In the case of semi-automatic annotation, there are still many open research and technical
|
|
issues. Even with perfect image analysis tools, a system cannot
|
|
recognise events that may have semantic meaning. This problem is due to the gap that exists
|
|
between low-level image analysis tools and high-level image annotations.
|
|
</p>
|
|
|
|
<h3>
|
|
<a name="solution_culture">
|
|
5.2 Use Case: Cultural Heritage
|
|
</a>
|
|
</h3>
|
|
|
|
<div style="float: right; width: 30%; border: 1px solid gray; padding: 1%; margin: 1%">
|
|
<img style="width: 100%;;"
|
|
src="http://www.artchive.com/artchive/m/monet/adresse.jpg"
|
|
alt="Image of Monet's painting 'Garden at Sainte-Adresse'"/>
|
|
|
|
Claude Monet, Garden at Sainte-Adresse.<br/>
|
|
|
|
Image courtesy of <a href="http://www.artchive.com">Mark
|
|
Harden</a>, used with permission.
|
|
</div>
|
|
|
|
<h4 id="ec_solution">Possible Semantic Web-based solution</h4>
|
|
|
|
<p>
|
|
Many of the requirements of the use case described in <a href="#culture_images">Section 2.2</a>
|
|
can be met by using the vocabulary developed by the <a href="#VraCore">VRA</a> in
|
|
combination with domain-specific vocabularies such as Getty's AAT
|
|
and ULAN.
|
|
|
|
In this section, we provide as an example a set of RDF annotations
|
|
of a painting by Claude Monet, which is in English known as "Garden
|
|
at Sainte-Adresse". It is part of the collection of the
|
|
Metropolitan Museum of Art in New York. The corresponding RDF file
|
|
is <a href="eculture-use-case.rdf">available as a
|
|
separate document</a>. No special annotation tools where used to
|
|
create the annotations. We assume that cultural heritage
|
|
organizations that need to publish similar metadata will do so by
|
|
exporting existing information from their collection database to
|
|
RDF. Below, we discuss the different annotations used in this
|
|
file.
|
|
</p>
|
|
|
|
|
|
<h4 id="housekeeping">House keeping</h4>
|
|
<p>
|
|
The file starts as a typical RDF/XML file, by defining the XML
|
|
version and encoding and defining entities for the RDF and VRA
|
|
namespaces that will be used later. Note that we use the <a
|
|
href="#VraRDF">RDF/OWL schema of VRA Core</a> developed by Mark
|
|
van Assem.
|
|
</p>
|
|
|
|
<div class="exampleInner" style="clear: both">
|
|
<pre>
|
|
<?xml version='1.0' encoding='ISO-8859-1'?>
|
|
<!DOCTYPE rdf:RDF [
|
|
<!ENTITY rdf "http://www.w3.org/1999/02/22-rdf-syntax-ns#">
|
|
<!ENTITY vra "http://www.vraweb.org/vracore/vracore3#">
|
|
</pre>
|
|
</div>
|
|
|
|
<h5 id="work_or_image">Work versus Image</h5>
|
|
<p>
|
|
The example includes annotations about two different images of the
|
|
same painting. An important distinction made by VRA vocabulary is
|
|
the distinction between annotations describing a work of art itself
|
|
and annotations describing (digital) images of that work. This
|
|
example also uses this distinction. In RDF, to say something about
|
|
a resource, that resource needs to have a URI. We will thus not
|
|
only need the URIs of the two images, but also a URI for the
|
|
painting itself:
|
|
</p>
|
|
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<!ENTITY image1 "http://www.metmuseum.org/Works_Of_Art/images/ep/images/ep67.241.L.jpg">
|
|
<!ENTITY image2 "http://www.artchive.com/artchive/m/monet/adresse.jpg">
|
|
<!ENTITY painting "http://thing-described-by.org/?http://www.metmuseum.org/Works_Of_Art/images/ep/images/ep67.241.L.jpg">
|
|
]>
|
|
</pre>
|
|
</div>
|
|
|
|
<h5 id="uri_conventions">URI and ID conventions</h5>
|
|
<p>
|
|
VRA Core does not specify how works, images or annotation records
|
|
should be identified. For the two images, we have chosen for the
|
|
most straightforward solution and use the URI of the image as the
|
|
identifying URI. We did not have, however, a similar URI that
|
|
identifies the painting itself. We could not reuse the URI of one
|
|
of the images. This is not only conceptually wrong, but would also
|
|
lead to technical errors: it would make the existing instance of
|
|
<tt>vra:Image</tt> also an instance of the <tt>vra:Work</tt> class,
|
|
while this is not allowed by the schema.
|
|
</p>
|
|
<p>
|
|
In the example, we have decided to `mint' the URI of the painting
|
|
by arbitrary selecting the URI of one of the images, and prefixing
|
|
it by <tt><a href="http://thing-described-by.org/">
|
|
http://thing-described-by.org/?</a></tt>. This creates a new URI
|
|
that is distinct from the image itself, but when a the browser
|
|
resolves it, it will be redirected to the image URI by the
|
|
<tt>thing-described-by.org</tt> web server (one could argue if the
|
|
use of an http-based URI is actually appropriate here. See <a
|
|
href="#HTTP-URI">What do HTTP URIs Identify?</a> and <a
|
|
href="#httpRange-14">[httpRange-14]</a> for more details on this
|
|
discussion).
|
|
</p>
|
|
<p>
|
|
Warning: The annotations described below also contain a
|
|
<tt>vra:idNumber.currentRepository</tt> element, that defines the
|
|
identifier used <em>locally</em> in the museum's repositories.
|
|
These local identifiers should not be confused with the globally
|
|
unique identifier that is provided by the URI.
|
|
</p>
|
|
|
|
<h5 id="housekeeping2">More housekeeping: starting the RDF block</h5>
|
|
<p>
|
|
The next line opens the RDF block, declares the namespaces using
|
|
the XML entities defined above. Out of courtesy, it uses
|
|
<tt>rdf:seeAlso</tt> to help agents find the VRA schema that is
|
|
used.
|
|
</p>
|
|
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<rdf:RDF xmlns:rdf="&rdf;" xmlns:vra="&vra;"
|
|
rdf:seeAlso="http://www.w3.org/2001/sw/BestPractices/MM/vracore3.rdfs"
|
|
>
|
|
</pre>
|
|
</div>
|
|
|
|
<h4 id="work">Description of the work (painting)</h4>
|
|
<p>
|
|
The following lines describe properties of the painting itself: we
|
|
will deal with the properties of the two images later. First, we
|
|
provide general information about the painting such as the title,
|
|
its creator and the date of creation. For these properties, the VRA
|
|
closely follows the Dublin Core conventions:
|
|
</p>
|
|
<div class="exampleInner">
|
|
<pre>
|
|
|
|
<!-- Description of the painting -->
|
|
<vra:Work rdf:about="&painting1;">
|
|
|
|
<!-- General information -->
|
|
<vra.title>Jardin à; Sainte-Adresse</vra.title>
|
|
<vra:title.translation>Garden at Sainte-Adresse</vra:title.translation>
|
|
<vra:creator>Monet, Claude</vra:creator> <!-- ULAN ID:500019484 -->
|
|
<vra:creator.role>artist</vra:creator.role> <!-- ULAN ID:31100 -->
|
|
<vra:date.creation>1867</vra:date.creation>
|
|
</pre>
|
|
</div>
|
|
|
|
<h5 id="text_or_controlled">Text fields and controlled vocabularies</h5>
|
|
<p>
|
|
Many values are filled with RDF Literals, of which the value is not
|
|
further constraint by the schema. But many of these values are
|
|
actually terms from other controlled vocabularies, such as the
|
|
Getty <a href="#refAAT">AAT</a>, <a href="#refULAN">ULAN</a> or a image
|
|
type defined by <a href="#refMIME-2">MIME</a>. Using controlled
|
|
vocabularies solves many problems associated with free text
|
|
annotations. For example, ULAN recommends a spelling when an
|
|
artist's name is used for indexing, so for the <tt>vra:creator</tt>
|
|
field we have exactly used this spelling ("Monet, Claude"). The
|
|
ULAN identifiers of the records describing Claude Monet and the
|
|
"artist" class are given in XML comments above. The use of
|
|
controlled vocabulary can avoid confusion and the need for
|
|
"smushing" different spellings for the same name later.
|
|
</p>
|
|
|
|
<p>
|
|
However, using controlled vocabularies does not solve the problem
|
|
of ambiguous terms. The annotations below use three different uses
|
|
of "oil paint", "oil paintings" and "oil painting (technique)".
|
|
The first refers to the type of paint used on the canvas, the
|
|
second to the type of work (e.g. the work is an oil painting, and
|
|
not an etching) and the last to the painting technique used by
|
|
artist. All three terms refer to different concepts that are part
|
|
of different branches of the AAT term hierarchy (the AAT
|
|
identifiers of these concepts are mentioned in XML comments).
|
|
However, the use of terms that are so similar for different
|
|
concepts is bound to lead to confusion. Instead, one could switch
|
|
from using <tt>owl:datatypeProperties</tt> to using
|
|
owl:objectProperties, and replace the literal text by a reference
|
|
to the URI of the concept used. For example, one could change:
|
|
|
|
<br/><tt><vra:material.medium>oil paint</vra:material.medium></tt>
|
|
<br/>to
|
|
<br/><tt><vra:material.medium rdf:resource="http://www.getty.edu/aat#300015050"/></tt>
|
|
</p>
|
|
|
|
<p>
|
|
This approach, requires, however, that an unambiguous URI-based
|
|
naming scheme is defined for all terms in the target vocabulary
|
|
(and in this case, such a URI-based naming scheme does not yet
|
|
exist for AAT terms). Additional Semantic Web-based processing is
|
|
also only possible once these vocabularies become available in RDF
|
|
or OWL.
|
|
</p>
|
|
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<!-- Technical information -->
|
|
<vra:measurements.dimensions>98.1 x 129.9 cm</vra:measurements.dimensions>
|
|
<vra:material.support>unprimed canvas</vra:material.support> <!-- AAT ID:300238097 -->
|
|
<vra:material.medium>oil paint</vra:material.medium> <!-- AAT ID:300015050 -->
|
|
<vra:type>oil paintings</vra:type> <!-- AAT ID:300033799 -->
|
|
<vra.technique>oil painting (technique)</vra.technique> <!-- AAT ID:300178684 -->
|
|
|
|
<!-- Associated style etc -->
|
|
<vra:stylePeriod>Impressionist</vra:stylePeriod> <!-- AAT ID:300021503 -->
|
|
<vra:culture>French</vra:culture> <!-- AAT ID:300111188 -->
|
|
</pre>
|
|
</div>
|
|
|
|
<h5 id="subject_matter">Annotating subject matter</h5>
|
|
<p>
|
|
For many applications, it is useful to know what is actually
|
|
depicted by the painting. One could add annotations of this style
|
|
to an arbitrary level of detail. To keep the example simple, we
|
|
have chosen to record only the names of the people that are
|
|
depicted on the painting, using the <tt>vra:subject</tt> field.
|
|
Also for simplicity, we have chosen not to annotate specific parts
|
|
or regions of the painting. This might have been appropriate, for
|
|
example, to identify the associated regions that depict the various
|
|
people in the painting:
|
|
</p>
|
|
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<!-- Subject matter: (who/what is depicted by this work -->
|
|
<vra:subject>Jeanne-Marguerite Lecadre (artist's cousin)</vra:subject>
|
|
<vra:subject>Madame Lecadre (artist's aunt)</vra:subject>
|
|
<vra:subject>Adolphe Monet (artist's father)</vra:subject>
|
|
</pre>
|
|
</div>
|
|
|
|
<h5 id="provenance">Provenance: annotating the past</h5>
|
|
<p>
|
|
Many of the fields below do not contain information about the
|
|
current situation of the painting, but information about places and
|
|
collections the painting has been in the past. This provides
|
|
provenance information that is important in this domain.
|
|
</p>
|
|
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<!-- Provenance -->
|
|
<vra:location.currentSite>Metropolitan Museum of Art, New York</vra:location.currentSite>
|
|
<vra:location.formerSite>Montpellier</vra:location.formerSite>
|
|
<vra:location.formerSite>Paris</vra:location.formerSite>
|
|
<vra:location.formerSite>New York</vra:location.formerSite>
|
|
<vra:location.formerSite>Bryn Athyn, Pa.</vra:location.formerSite>
|
|
<vra:location.formerSite>London</vra:location.formerSite>
|
|
<vra:location.formerRepository>
|
|
Victor Frat, Montpellier (probably before 1870 at least 1879;
|
|
bought from the artist); his widow, Mme Frat, Montpellier (until 1913)
|
|
</vra:location.formerRepository>
|
|
<vra:location.formerRepository>Durand-Ruel, Paris, 1913</vra:location.formerRepository>
|
|
<vra:location.formerRepository>Durand-Ruel, New York, 1913</vra:location.formerRepository>
|
|
<vra:location.formerRepository>
|
|
Reverend Theodore Pitcairn and the Beneficia Foundation, Bryn Athyn, Pa. (1926-1967),
|
|
sale, Christie's, London, December 1, 1967, no. 26 to MMA
|
|
</vra:location.formerRepository>
|
|
<vra:idNumber.currentRepository>67.241</vra:idNumber.currentRepository> <!-- MMA ID number -->
|
|
</pre>
|
|
</div>
|
|
|
|
<h5 id="copyright">Copyright and origin of metadata</h5>
|
|
<p>
|
|
The remaining properties describe the origin the sources used for
|
|
creating the metadata and a rights management statement. We have
|
|
used the <tt>vra:description</tt> element to provide a link to a
|
|
web page with additional descriptive information:
|
|
</p>
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<!-- extra information, source of this information and copyright issues: -->
|
|
<vra:description>For more information, see http://www.metmuseum.org/Works_Of_Art/viewOne.asp?dep=11&viewmode=1&item=67%2E241&section=description#a</vra:description>
|
|
<vra:source>Metropolitan Museum of Art, New York</vra:source>
|
|
<vra:rights>Metropolitan Museum of Art, New York</vra:rights>
|
|
</pre>
|
|
</div>
|
|
|
|
<h4 id="images">Image properties</h4>
|
|
<p>
|
|
Finally, we define the properties that are specific to the two
|
|
images of the painting, which differ in resolution, copyright etc.
|
|
The first set of annotations describe a 500x300 pixel image that is
|
|
located at the website of the Metropolitan itself, while the second
|
|
set describes the properties of a larger resolution (1075 x 778px)
|
|
image at Mark Harden's <a
|
|
href="http://www.artchive.com/">Artchive</a> website.
|
|
Note that VRA Core does not specify how Works and their associated Images
|
|
should be related. In the example we follow <a href="#VraRDF">Van
|
|
Assem's suggestion</a> and use <tt>vra.relation.depicts</tt> to
|
|
explicitly link the Image to the Work it depicts.
|
|
</p>
|
|
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<!-- Description of the first online image of the painting -->
|
|
<vra:Image rdf:about="&image1a;">
|
|
<vra:type>digital images</vra:type> <!-- AAT ID: 300215302 -->
|
|
<vra:relation.depicts rdf:resource="&painting1;"/>
|
|
<vra.measurements.format>image/jpeg</vra.measurements.format> <!-- MIME -->
|
|
<vra.measurements.resolution>500 x 380px</vra.measurements.resolution>
|
|
<vra.technique>Scanning</vra.technique>
|
|
<vra:creator>Anonymous employee of the museum</vra:creator>
|
|
<vra:idNumber.currentRepository>ep67.241.L.jpg</vra:idNumber.currentRepository>
|
|
<vra:rights>Metropolitan Museum of Art, New York</vra:rights>
|
|
</vra:Image>
|
|
</pre>
|
|
</div>
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<!-- Description of the second online image of the painting -->
|
|
<vra:Image rdf:about="&image1b;">
|
|
<vra:type>digital images</vra:type> <!-- AAT ID: 300215302 -->
|
|
<vra:relation.depicts rdf:resource="&painting1;"/>
|
|
<vra:creator>Mark Harden</vra:creator>
|
|
<vra.technique>Scanning</vra.technique>
|
|
<vra.measurements.format>image/jpeg</vra.measurements.format> <!-- MIME -->
|
|
<vra.measurements.resolution>1075 x 778px</vra.measurements.resolution>
|
|
<vra:idNumber.currentRepository>adresse.jpg</vra:idNumber.currentRepository>
|
|
<vra:rights>Mark Harden, The Artchive, http://www.artchive.com/</vra:rights>
|
|
</vra:Image>
|
|
</rdf:RDF>
|
|
</pre>
|
|
</div>
|
|
|
|
<h4 id="ch_conclusion">Conclusion and discussion</h4>
|
|
<p>
|
|
The example above reveals several technical issues that are still
|
|
open. For example, the way the URI for the painting was minted is
|
|
rather arbitrary. Preferably, there would have been a commonly
|
|
accepted URI scheme for paintings (c.f. the <a href="#lsid">LSID</a>
|
|
scheme used to identify concepts from the life sciences). At the
|
|
time of writing, the VRA, AAT and ULAN vocabulary used have
|
|
currently no commonly agreed upon RDF or OWL representation, which
|
|
reduces the interoperability of the chosen approach. Tool support
|
|
is another issue. While some major database vendors already start
|
|
to support RDF, generating the type of RDF as shown here from
|
|
existing collection databases will in many cases require non
|
|
trivial custom conversion software.
|
|
</p>
|
|
<p>
|
|
From a modeling point of view, subject matter annotations are
|
|
always non-trivial. As stated above, it is hard to give general
|
|
guidelines about what should be annotated and to what depth, as
|
|
this can be very application dependent. Note that in the example,
|
|
we annotated the persons that appear in the painting, and that we
|
|
modeled this information as properties of the painting URI, not of
|
|
the two image URIs. But if we slightly modify our use case and
|
|
assume one normal image and one X-ray image that reveals an older
|
|
painting under this one, it might make more sense to model more
|
|
specific subject matter annotations as properties of the specific
|
|
images.
|
|
</p>
|
|
|
|
<p>
|
|
Nevertheless, the example shows that a large part of issues
|
|
described by the use case can be solved using current Semantic Web
|
|
technology. It shows how RDF can be used to use existing
|
|
vocabularies to annotate various aspects of paintings and the
|
|
images that depict them.
|
|
</p>
|
|
|
|
<h3>
|
|
<a name="solution_TVarchive">
|
|
5.3 Use Case: Television News Archive
|
|
</a>
|
|
</h3>
|
|
|
|
<h4 id="archive_solution">Possible Semantic Web-based solution</h4>
|
|
|
|
<p>
|
|
The use case described in <a href="#media_images">Section 2.3</a>
|
|
is typically one that requires the use of multiple
|
|
vocabularies. Let us imagine that the image to be described is about
|
|
a refused goal of a given soccer player (e.g. J.A Boumsong) for
|
|
an active offside position during a particular game (e.g. Auxerre-Metz).
|
|
First, the image can be extracted from a weekly sports
|
|
magazine broadcasted on a TV channel. This program may be fully
|
|
described using the vocabulary developed by the <a href="#TVA">
|
|
[TV Anytime forum]</a>. Second, this image shows the player
|
|
Jean-Alain Boumsong scoring with his head during the game
|
|
Auxerre-Metz. The context of this football game could be described
|
|
using the <a href="#MPEG-7">[MPEG-7]</a> vocabulary while the
|
|
action itself might be described by a soccer ontology such as the
|
|
one developed by <a href="#Tsinaraki">[Tsinaraki]</a>. Finally, a
|
|
soccer fan may notice that this goal was actually refused for an
|
|
active offside position of another player. On the image, a circle
|
|
could highlight this player badly positioned. Again, the description
|
|
could merge MPEG-7 vocabulary for delimiting the relevant image
|
|
region and a domain specific ontology for describing the action
|
|
itself.
|
|
|
|
In the following, we provide as an example a set of RDF annotations
|
|
illustrating these three levels of description as well as the
|
|
vocabularies involved.
|
|
</p>
|
|
|
|
<h5>The image context</h5>
|
|
<p>
|
|
Let us consider that the image comes from a weekly sports magazine named <a
|
|
href="http://sport.france2.fr/stade2/">Stade 2</a> broadcasted on
|
|
March, 17th 2002 on the French public channel <a
|
|
href="http://www.france2.fr/">France 2</a>. This context can be
|
|
represented using the TV Anytime vocabulary which allows for a TV
|
|
(or radio) broadcaster to publish its program listings on the web
|
|
or in an electronic program guide. Therefore, this vocabulary
|
|
provides the necessary concepts and relations for cataloging the
|
|
programs, giving their intended audience, format and genre, or some
|
|
parental guidance. The vocabulary contains also the vocabulary for
|
|
describing afterwards the real audience and the peak viewing times
|
|
which are of crucial importance for the broadcasters in order to
|
|
adapt their advertisement rates.
|
|
</p>
|
|
|
|
<div style="clear: both;" class="exampleOuter">
|
|
<div class="c1">
|
|
<a id="exampleTV" name="exampleTV">RDF description of the program from which the image comes from</a>
|
|
</div>
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<?xml version='1.0' encoding='ISO-8859-1'?>
|
|
<!DOCTYPE rdf:RDF [
|
|
<!ENTITY rdf "http://www.w3.org/1999/02/22-rdf-syntax-ns#">
|
|
<!ENTITY xsd "http://www.w3.org/2001/XMLSchema#">
|
|
]>
|
|
|
|
<rdf:RDF
|
|
xmlns:rdf="&rdf;"
|
|
xmlns:xsd="&xsd;"
|
|
xmlns:tva="urn:tva:metadata:2002"
|
|
>
|
|
|
|
<tva:Program rdf:about="program1">
|
|
<tva:hasTitle>Stade 2</tva:hasTitle>
|
|
<tva:hasSynopsis>Weekly Sports Magazine broadcasted every Sunday</tva:hasSynopsis>
|
|
<tva:Genre rdf:resource="urn:tva:metadata:cs:IntentionCS:2002:Entertainment"/>
|
|
<tva:Genre rdf:resource="urn:tva:metadata:cs:FormatCS:2002:Magazine"/>
|
|
<tva:Genre rdf:resource="urn:tva:metadata:cs:ContentCS:2002:Sports"/>
|
|
<tva:ReleaseInformation>
|
|
<rdf:Description>
|
|
<tva:ReleaseDate xsd:date="2002-03-17"/>
|
|
<tva:ReleaseLocation>fr</tva:ReleaseLocation>
|
|
</rdf:Description>
|
|
</tva:ReleaseInformation>
|
|
</tva:Program>
|
|
|
|
</rdf:RDF>
|
|
</pre>
|
|
</div>
|
|
</div>
|
|
|
|
<h5>The description of the action</h5>
|
|
<p>
|
|
To be done.
|
|
</p>
|
|
|
|
<h5>The description of particular region</h5>
|
|
<p>
|
|
Discuss the pros and cons of having either 2 separate files (one
|
|
expressing the localization of the region and one representing the
|
|
content annotation) or 1 RDF file having both description.
|
|
</p>
|
|
|
|
<h5>The annotation link</h5>
|
|
<p>
|
|
Discuss the various annotation links provided by MPEG-7 (annotates, depicts, exemplifies, etc).
|
|
</p>
|
|
|
|
|
|
<h3>
|
|
<a name="solution_NASA">
|
|
5.4 Use Case: large-scale image collections at NASA
|
|
</a>
|
|
</h3>
|
|
|
|
<div style="float: right; width: 317px; height:540px; border: 1px solid gray; padding: 1%; margin: 1%">
|
|
<a href="http://grin.hq.nasa.gov/IMAGES/SMALL/GPN-2000-001171.jpg">
|
|
<img style="width: 317px;height:450px;"
|
|
src="http://grin.hq.nasa.gov/IMAGES/SMALL/GPN-2000-001171.jpg"
|
|
alt="Apollo 7 Saturn rocket launch"/></a>
|
|
<br/>
|
|
Apollo 7 Saturn rocket launch -
|
|
October, 10th 1968. Image courtesy of NASA, available at <a href="http://grin.hq.nasa.gov/">GRIN</a>,
|
|
used with permission.
|
|
</div>
|
|
|
|
<h4 id="large-collection-solution">Possible Semantic Web-based solution</h4>
|
|
<p>
|
|
One possible solution for the requirements expressed in the use case description
|
|
in <a href="#scientific_images">Section 2.4</a>
|
|
is an annotation environment that enables users to annotate information
|
|
about images and/or their regions using concepts in ontologies
|
|
(OWL and/or RDFS). More specifically, subject matter experts will
|
|
be able to assert metadata elements about images and their
|
|
specific content. Multimedia related ontologies can be used to
|
|
localize and represent regions within particular images. These
|
|
regions can then be related to the image via a
|
|
depiction/annotation property. This functionality can be provided,
|
|
for example, by the <a
|
|
href="http://www.mindswap.org/2005/owl/digital-media">MINDSWAP
|
|
digital-media ontology</a> (to represent images, image regions,
|
|
etc.), in conjunction with <a
|
|
href="http://xmlns.com/foaf/0.1/">FOAF</a> (to assert image
|
|
depictions). Additionally, in order to represent the low level
|
|
image features of regions, the <a
|
|
href="http://www.acemedia.org/aceMedia/reference/resource/index.html">aceMedia
|
|
Visual Descriptor Ontology</a> can be used.
|
|
</p>
|
|
|
|
<h5>Domain Specific Ontologies</h5>
|
|
|
|
<p>
|
|
In order to describe the content of such images, a mechanism to
|
|
represent the domain specific content depicted within them is
|
|
needed. For this use case, domain ontologies that define space
|
|
specific concepts and relations can be used. Such ontologies are
|
|
freely available and include, but are not limited to the following:
|
|
</p>
|
|
<ul>
|
|
<li> <a href="http://semspace.mindswap.org/2004/ontologies/ShuttleMission-ont.owl">Shuttle related (OWL)</a></li>
|
|
<li> <a href="http://semspace.mindswap.org/2004/ontologies/ShuttleMission-ont.rdfs">Shuttle related (RDFS)</a></li>
|
|
<li> <a href="http://semspace.mindswap.org/2004/ontologies/System-ont.owl">Space vehicle system related (OWL)</a></li>
|
|
<li> <a href="http://semspace.mindswap.org/2004/ontologies/System-ont.rdfs">Space vehicle system related (RDFS)</a></li>
|
|
</ul>
|
|
|
|
<h5>Visual Ontologies</h5>
|
|
<p>
|
|
As discussed above, this scenario requires the ability to state
|
|
that images (and possibly their regions) depict certain things. For
|
|
example, consider a picture of the <a
|
|
href="http://grin.hq.nasa.gov/IMAGES/SMALL/GPN-2000-001171.jpg">Apollo
|
|
7 Saturn rocket launch</a>. One would want to make assertions that
|
|
include that the image <i>depicts</i> the Apollo 7 launch, the
|
|
Apollo 7 Saturn IB space vehicle is depicted in a rectangular
|
|
<i>region</i> around the rocket, the image <i>creator</i> is NASA,
|
|
etc. One possible way to accomplish this is to use a combination of
|
|
various multimedia related ontologies, including <a
|
|
href="http://xmlns.com/foaf/0.1/">FOAF</a> and the <a
|
|
href="http://www.mindswap.org/2005/owl/digital-media">MINDSWAP
|
|
digital-media ontology</a>. More specifically, image depictions can
|
|
be asserted via a <i>depiction</i> property (a sub-property of
|
|
foaf:depiction) defined in the MINDSWAP Digital Media
|
|
ontology. Thus, images can be semantically linked to instances
|
|
defined on the Web. Image regions can defined via an
|
|
<i>ImagePart</i> concept (also defined in the MINDSWAP Digital
|
|
Media ontology). Additionally, regions can be given a bounding box
|
|
by using a property named <i>svgOutline</i>, allowing localizing of
|
|
image parts. Essentially SVG outlines (SVG XML literals) of the
|
|
regions can be specified using this property. Using the <a
|
|
href="http://dublincore.org/schemas/rdfs//">Dublin Core
|
|
standard</a> and the <a href="http://www.w3.org/2003/12/exif/">EXIF
|
|
Schema </a> more general annotations about the image can be stated
|
|
as well, including its creator, size, etc. A subset of these sample
|
|
annotations are shown in an RDF graph below in <a
|
|
href="#figure2">Figure 2</a>.
|
|
</p>
|
|
|
|
<div style="text-align: center">
|
|
<img alt="RDF Graph Describing the Apollo 7 Launch Image"
|
|
style="width: 100%;"
|
|
src="nasaRDFDiagram.png"/>
|
|
<br/>
|
|
<a id="figure2" name="figure2">Figure 2: An RDF Graph
|
|
Describing the Apollo 7 Launch Image</a>
|
|
</div>
|
|
|
|
<p>
|
|
<a href="#figure2">Figure 2</a> illustrates how the approach links metadata to the image:</p>
|
|
|
|
<ul>
|
|
<li>image content, e.g., Apollo 7 Launch, is identified by
|
|
<code>http://www.mindswap.org/2005/owl/digital-media#depicts</code></li>
|
|
|
|
<li>image subparts are identified by the property
|
|
<code>http://www.mindswap.org/2005/owl/digital-media#hasRegion</code></li>
|
|
|
|
<li>image regions are localized using
|
|
<code>http://www.mindswap.org/2005/owl/digital-media#svgOutline</code> and an SVG snippet</li>
|
|
|
|
</ul>
|
|
|
|
|
|
|
|
<p>
|
|
Additionally, the entire annotations of the Apollo 7 launch are shown below in RDF/XML.<br/><br/>
|
|
</p>
|
|
<div class="c1">
|
|
<a id="exampleApollo7" name="exampleApollo7">RDF/XML annotations of Apollo 7 launch </a>
|
|
</div>
|
|
|
|
<div class="exampleInner">
|
|
<pre>
|
|
<rdf:RDF
|
|
xmlns:j.0="http://www.w3.org/2003/12/exif/ns#"
|
|
xmlns:j.1="http://www.mindswap.org/2005/owl/digital-media#"
|
|
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
|
|
xmlns:j.2="http://semspace.mindswap.org/2004/ontologies/System-ont.owl#"
|
|
xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
|
|
xmlns:owl="http://www.w3.org/2002/07/owl#"
|
|
xmlns:dc="http://purl.org/dc/elements/1.1/"
|
|
xmlns:j.3="http://semspace.mindswap.org/2004/ontologies/ShuttleMission-ont.owl#"
|
|
xml:base="http://example.org/NASA-Use-Case" >
|
|
|
|
<rdf:Description rdf:about="A0">
|
|
<j.1:depicts rdf:resource="#Saturn_1B"/>
|
|
<rdf:type rdf:resource="http://www.mindswap.org/~glapizco/technical.owl#ImagePart"/>
|
|
<rdfs:label>region2407</rdfs:label>
|
|
<j.1:regionOf rdf:resource="http://grin.hq.nasa.gov/IMAGES/SMALL/GPN-2000-001171.jpg"/>
|
|
<j.1:svgOutline>
|
|
<svg xml:space="preserve" width="451" heigth="640" viewBox="0 0 451 640">
|
|
<image xlink:href="http://grin.hq.nasa.gov/IMAGES/SMALL/GPN-2000-001171.jpg" x="0" y="0" width="451" height="640" />
|
|
<rect x="242.0" y="79.0" width="46.0" height="236.0" style="fill:none; stroke:yellow; stroke-width:1pt;"/>
|
|
</svg>
|
|
</j.1:svgOutline>
|
|
</rdf:Description>
|
|
|
|
<rdf:Description rdf:about="http://grin.hq.nasa.gov/IMAGES/SMALL/GPN-2000-001171.jpg">
|
|
<j.0:imageLength>640</j.0:imageLength>
|
|
<dc:date>10/11/1968</dc:date>
|
|
<dc:description>Taken at Kennedy Space Center in Florida</dc:description>
|
|
<j.1:depicts rdf:resource="#Apollo_7_Launch"/>
|
|
<j.1:hasRegion rdf:nodeID="A0"/>
|
|
<dc:creator>NASA</dc:creator>
|
|
<rdf:type rdf:resource="http://www.mindswap.org/~glapizco/technical.owl#Image"/>
|
|
<j.0:imageWidth>451</j.0:imageWidth>
|
|
</rdf:Description>
|
|
|
|
<rdf:Description rdf:about="#Apollo_7_Launch">
|
|
<j.3:launchDate>10/11/1968</j.3:launchDate>
|
|
<j.3:codeName>Apollo 7 Launch</j.3:codeName>
|
|
<j.3:has_shuttle rdf:resource="#Saturn_1B"/>
|
|
|
|
<rdfs:label>Apollo 7 Launch</rdfs:label>
|
|
<j.1:depiction rdf:resource="http://grin.hq.nasa.gov/IMAGES/SMALL/GPN-2000-001171.jpg"/>
|
|
<rdf:type rdf:resource="http://semspace.mindswap.org/2004/ontologies/ShuttleMission-ont.owl#Launch"/>
|
|
</rdf:Description>
|
|
|
|
<rdf:Description rdf:about="#Saturn_1B">
|
|
<rdfs:label>Saturn_1B</rdfs:label>
|
|
<j.1:depiction rdf:nodeID="A1"/>
|
|
<rdfs:label>Saturn 1B</rdfs:label>
|
|
<rdf:type rdf:resource="http://semspace.mindswap.org/2004/ontologies/System-ont.owl#ShuttleName"/>
|
|
<j.1:depiction rdf:nodeID="A0"/>
|
|
</rdf:Description>
|
|
|
|
</rdf:RDF>
|
|
</pre>
|
|
</div>
|
|
|
|
|
|
<p>
|
|
In order to represent the low level features of images, the <a
|
|
href="http://www.acemedia.org/aceMedia/reference/resource/index.html">aceMedia
|
|
Visual Descriptor Ontology</a> can be used. This ontology contains
|
|
representations of MPEG-7 visual descriptors and models Concepts
|
|
and Properties that describe visual characteristics of objects. For
|
|
example, the dominant color descriptor can be used to describe the
|
|
number and value of dominant colors that are present in a region of
|
|
interest and the percentage of pixels that each associated color
|
|
value has.
|
|
</p>
|
|
|
|
|
|
<h5>Available Annotation Tools</h5>
|
|
<p>
|
|
Existing toolkits, such as <a href="#PS">[PhotoStuff]</a> and <a
|
|
href="#mOnto">[M-OntoMat-Annotizer]</a>, currently provide
|
|
graphical environments to accomplish the annotation tasks mentioned
|
|
above. Using such tools, users can load images, create regions
|
|
around parts of the image, automatically extract low-level features
|
|
of selected regions (via M-OntoMat-Annotizer), assert statements
|
|
about the selected regions, etc. Additionally, the resulting
|
|
annotations can be exported as RDF/XML (as shown above), thus
|
|
allowing them be shared, indexed, and used by advanced
|
|
annotation-based browsing (and searchable) environments.
|
|
</p>
|
|
|
|
<!-- ===================================================================== -->
|
|
<h2>
|
|
<a name="conclusions" id="conclusions">
|
|
6. Conclusions
|
|
</a>
|
|
</h2>
|
|
<p>
|
|
Current Semantic Web technologies are sufficiently generic to support
|
|
annotation of a wide variety of Web resources, including image
|
|
resources. This document provides examples of the use of Semantic
|
|
Web languages and tools for image annotation, based on use cases
|
|
for a wide variety of domains. It also briefly surveys some
|
|
currently available vocabularies and tools that can be used to
|
|
semantically annotate images so that machine can better process them.
|
|
The use of Semantic Web technologies have significant advantages in applications areas in
|
|
which the interoperability of heterogeneous metadata is important
|
|
and in areas that require an explicitly defined and formal
|
|
semantics of the metadata in order to perform reasoning tasks.
|
|
</p>
|
|
|
|
<p>
|
|
Still, many things need to be improved. Commonly accepted, widely
|
|
used vocabularies for image annotation are still missing. Having
|
|
such vocabularies would help in sharing metadata across
|
|
applications and across multiple domains. Especially, a standard
|
|
means to address subregions withing an image is still missing. In
|
|
addition, tool support needs to improve dramatically before
|
|
Semantic Web-based image annotation can be applied on an industrial
|
|
scale: support needs to be integrated in the entire production and
|
|
distribution chain. Finally, many existing approaches for image
|
|
metadata are not based on Semantic Web technologies, and work is
|
|
required to make these approaches interoperable with the Semantic Web.
|
|
</p>
|
|
|
|
<!-- ===================================================================== -->
|
|
<h2>
|
|
<a name="references" id="references">
|
|
References</a>
|
|
</h2>
|
|
|
|
<dl>
|
|
|
|
<dt>
|
|
<a id="refAAT" name="refAAT">[AAT]</a>
|
|
</dt>
|
|
<dd><span class="title">Art and Architecture Thesaurus</span>.
|
|
The J. Paul Getty Trust, 2004.
|
|
(See <a href="http://www.getty.edu/research/conducting_research/vocabularies/aat/">
|
|
http://www.getty.edu/research/conducting_research/vocabularies/aat/</a>)
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="DublinCore" name="DublinCore">[Dublin Core]</a>
|
|
</dt>
|
|
<dd>
|
|
The Dublin Core Metadata Initiative,
|
|
<a href="http://dublincore.org/documents/dces/">Dublin Core Metadata Element Set, Version 1.1: Reference Description</a>.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="httpRange-14" name="httpRange-14">
|
|
[httpRange-14]</a></dt>
|
|
<dd>
|
|
TAG's issue list, issue 14, see
|
|
<a href="http://www.w3.org/2001/tag/issues.html?type=1#httpRange-14">
|
|
http://www.w3.org/2001/tag/issues.html?type=1#httpRange-14</a>
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="HTTP-URI" name="HTTP-URI">
|
|
[HTTP-URI]</a></dt>
|
|
<dd>
|
|
Tim Berners-Lee, What do HTTP URIs Identify? Available at <a
|
|
href="http://www.w3.org/DesignIssues/HTTP-URI">http://www.w3.org/DesignIssues/HTTP-URI</a>
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="Hunter01" name="Hunter01">[Hunter, 2001]</a>
|
|
</dt>
|
|
<dd>
|
|
J. Hunter.
|
|
<!-- official link broken? a href="http://www.semanticweb.org/SWWS/program/full/paper59.pdf" -->
|
|
<a href="http://archive.dstc.edu.au/RDU/staff/jane-hunter/swws.pdf">Adding
|
|
Multimedia to the Semantic Web — Building an MPEG-7
|
|
Ontology</a>. In <i><a
|
|
href="http://www.semanticweb.org/SWWS/">International Semantic Web
|
|
Working Symposium (SWWS 2001)</a></i>, Stanford University,
|
|
California, USA, July 30 - August 1, 2001.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="lsid" name="lsid">
|
|
[LSID]
|
|
</a>
|
|
</dt>
|
|
<dd>
|
|
Life Sciences Identifier specification, <a
|
|
href="http://www.omg.org/cgi-bin/doc?dtc/04-05-01">
|
|
http://www.omg.org/cgi-bin/doc?dtc/04-05-01</a>.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a name="refMIME-2" id="refMIME-2">[MIME-2]</a>
|
|
</dt>
|
|
<dd>
|
|
<a href="ftp://ftp.isi.edu/in-notes/rfc2046.txt">
|
|
RFC 2046: Multipurpose Internet Mail Extensions (MIME) Part Two:
|
|
Media Types
|
|
</a>
|
|
. N. Freed, N. Borenstein, November 1996. Available at
|
|
<a href="ftp://ftp.isi.edu/in-notes/rfc2046.txt">ftp://ftp.isi.edu/in-notes/rfc2046.txt</a>
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="mOnto" name="mOnto">[M-OntoMat-Annotizer]</a>
|
|
</dt>
|
|
<dd>
|
|
M-OntoMat-AnnotizerProject Homepage at <a
|
|
href="http://www.acemedia.org/aceMedia/results/software/m-ontomat-annotizer.html">
|
|
http://www.acemedia.org/aceMedia/results/software/m-ontomat-annotizer.html</a>
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="MPEG-7" name="MPEG-7">[MPEG-7]</a>
|
|
</dt>
|
|
<dd>
|
|
Information Technology - Multimedia Content Description Interface (MPEG-7).
|
|
Standard No. ISO/IEC 15938:2001, International Organization for Standardization(ISO), 2001.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="Ossenbruggen04" name="Ossenbruggen04">[Ossenbruggen, 2004]</a>
|
|
</dt>
|
|
<dd>
|
|
J. van Ossenbruggen, F. Nack, and L. Hardman. That Obscure Object of Desire: Multimedia Metadata on the Web (Part I). In:
|
|
IEEE Multimedia 11(4), pp. 38-48 October-December 2004.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="Ossenbruggen05" name="Ossenbruggen05">[Ossenbruggen, 2005]</a>
|
|
</dt>
|
|
<dd>
|
|
F. Nack, J. van Ossenbruggen, and L. Hardman. That Obscure Object of Desire: Multimedia Metadata on the Web (Part II). In:
|
|
IEEE Multimedia 12(1), pp. 54-63 January-March 2005.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a name="OWL-Guide" id="OWL-Guide">[OWL Guide]</a>
|
|
</dt>
|
|
<dd>
|
|
<cite>
|
|
<a href="http://www.w3.org/TR/2004/REC-owl-guide-20040210/">
|
|
OWL Web Ontology Language Guide</a></cite>, Michael K.
|
|
Smith, Chris Welty, and Deborah L. McGuinness, Editors, W3C
|
|
Recommendation, 10 February 2004,
|
|
http://www.w3.org/TR/2004/REC-owl-guide-20040210/ .
|
|
<a href="http://www.w3.org/TR/owl-guide/">Latest
|
|
version</a> available at http://www.w3.org/TR/owl-guide/ .
|
|
</dd>
|
|
|
|
<dt>
|
|
<a name="OWL" id="OWL">[OWL Semantics and Abstract Syntax]</a></dt>
|
|
<dd>
|
|
<cite>
|
|
<a href=
|
|
"http://www.w3.org/TR/2004/REC-owl-semantics-20040210/">OWL Web
|
|
Ontology Language Semantics and Abstract Syntax</a></cite>, Peter
|
|
F. Patel-Schneider, Patrick Hayes, and Ian Horrocks, Editors, W3C
|
|
Recommendation 10 February 2004,
|
|
http://www.w3.org/TR/2004/REC-owl-semantics-20040210/ . <a href=
|
|
"http://www.w3.org/TR/owl-semantics/">Latest version</a>
|
|
available at http://www.w3.org/TR/owl-semantics/ .</dd>
|
|
|
|
<dt>
|
|
<a id="PS" name="PS">[PhotoStuff]</a>
|
|
</dt>
|
|
<dd>
|
|
PhotoStuff Project Homepage at <a
|
|
href="http://www.mindswap.org/2003/PhotoStuff/">http://www.mindswap.org/2003/PhotoStuff/</a>
|
|
</dd>
|
|
|
|
<dt><a id="RDF-Primer" name="RDF-Primer">[RDF Primer]</a></dt>
|
|
<dd>
|
|
<cite><a href="http://www.w3.org/TR/2004/REC-rdf-primer-20040210/">RDF
|
|
Primer</a></cite>, F. Manola, E. Miller, Editors, W3C Recommendation, 10 February 2004. <a href="http://www.w3.org/TR/2004/REC-rdf-primer-20040210/">This
|
|
version</a> is
|
|
http://www.w3.org/TR/2004/REC-rdf-primer-20040210/. The <a href="http://www.w3.org/TR/rdf-primer/">latest version</a> is at
|
|
http://www.w3.org/TR/rdf-primer/.
|
|
</dd>
|
|
|
|
<dt><a id="RDF" name="RDF"></a>[RDF Syntax]</dt>
|
|
<dd>
|
|
<cite>
|
|
<a href="http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/">
|
|
RDF/XML Syntax Specification (Revised)</a></cite>, Dave Beckett,
|
|
Editor, W3C Recommendation, 10 February 2004,
|
|
http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/ . <a
|
|
href="http://www.w3.org/TR/rdf-syntax-grammar/">Latest
|
|
version</a> available at
|
|
http://www.w3.org/TR/rdf-syntax-grammar/ .</dd>
|
|
|
|
|
|
<dt>
|
|
<a id="Stamou05" name="Stamou05">[Stamou, 2005]</a>
|
|
</dt>
|
|
<dd>
|
|
G. Stamou and S. Kollias (eds). Multimedia Content and the
|
|
Semantic Web: Methods, Standards and Tools. John Wiley & Sons
|
|
Ltd, 2005.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="Troncy03" name="Troncy03">[Troncy, 2003]</a>
|
|
</dt>
|
|
<dd>
|
|
R. Troncy. <a
|
|
href="http://springerlink.metapress.com/openurl.asp?genre=article&issn=0302-9743&volume=2870&spage=566">
|
|
Integrating Structure and Semantics into Audio-visual
|
|
Documents</a>. In <i><a
|
|
href="http://iswc2003.semanticweb.org/">Second International
|
|
Semantic Web Conference (ISWC 2003)</a></i>, pages 566 –
|
|
581, Sanibel Island, Florida, USA, October 20-23,
|
|
2003. Springer-Verlag Heidelberg.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="Tsinaraki" name="Tsinaraki">
|
|
[Tsinaraki]</a></dt>
|
|
<dd>
|
|
Tsinaraki, C.: OWL soccer ontology available at
|
|
<a href="http://elikonas.ced.tuc.gr/ontologies/soccer.zip">http://elikonas.ced.tuc.gr/ontologies/soccer.zip</a>.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="TVA" name="TVA">[TV Anytime]</a>
|
|
</dt>
|
|
<dd>
|
|
TV Anytime Forum,
|
|
<a href="http://www.tv-anytime.org/">
|
|
http://www.tv-anytime.org/
|
|
</a>
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="refULAN" name="refULAN">[ULAN]</a>
|
|
</dt>
|
|
<dd><span class="title">Union List of Artist Names</span>.
|
|
The J. Paul Getty Trust, 2004.
|
|
(See <a href="http://www.getty.edu/research/conducting_research/vocabularies/ulan/">
|
|
http://www.getty.edu/research/conducting_research/vocabularies/ulan/</a>)
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="VDO" name="VDO">[VDO]</a>
|
|
</dt>
|
|
<dd>
|
|
aceMedia Visual Descriptor Ontology, available from <a
|
|
href="http://www.acemedia.org/aceMedia/reference/resource/index.html">
|
|
http://www.acemedia.org/aceMedia/reference/resource/index.html</a>
|
|
</dd>
|
|
|
|
|
|
<dt>
|
|
<a id="VraCore" name="VraCore">[VRA Core]</a>
|
|
</dt>
|
|
<dd>
|
|
Visual Resources Association Data Standards Committee,
|
|
<a href="http://www.vraweb.org/vracore3.htm">
|
|
VRA Core Categories, Version 3.0</a>. Available at:
|
|
<a href="http://www.vraweb.org/vracore3.htm">
|
|
http://www.vraweb.org/vracore3.htm</a>.
|
|
</dd>
|
|
|
|
|
|
<dt>
|
|
<a id="VraRDF" name="VraRDF">[VRA in RDF/OWL]</a>
|
|
</dt>
|
|
|
|
<dd>
|
|
Mark van Assem. <a href="http://www.w3.org/2001/sw/BestPractices/MM/vra-conversion.html">
|
|
http://www.w3.org/2001/sw/BestPractices/MM/vra-conversion.html</a>
|
|
describes the RDFS schema of VRA Core 3.0 used in
|
|
<a href="#solution_culture">section 5.2</a>.
|
|
</dd>
|
|
|
|
<dt>
|
|
<a id="XML-NS" name="XML-NS">[XML NS]</a>
|
|
</dt>
|
|
<dd>
|
|
<cite>
|
|
<a href="http://www.w3.org/TR/1999/REC-xml-names-19990114/">Namespaces
|
|
in XML</a></cite>, Bray T., Hollander D., Layman A.
|
|
(Editors), World Wide Web Consortium, 14 January 1999. <a href="http://www.w3.org/TR/1999/REC-xml-names-19990114/">This
|
|
version</a> is http://www.w3.org/TR/1999/REC-xml-names-19990114/. The <a href="http://www.w3.org/TR/REC-xml-names/">latest version</a>
|
|
is http://www.w3.org/TR/REC-xml-names/.
|
|
</dd>
|
|
|
|
</dl>
|
|
|
|
<!-- ======================================================================== -->
|
|
|
|
<h2>
|
|
<a id="acknowledgments" name="acknowledgments">Acknowledgments</a>
|
|
</h2>
|
|
|
|
<p>
|
|
The editors would like to thank
|
|
John Smith (IBM T. J. Watson Research Center), Chris Catton (University of Oxford)
|
|
and the following Working Group members
|
|
for their feedback on earlier versions of this document:
|
|
Mark van Assem,
|
|
Jeremy Caroll,
|
|
Jane Hunter,
|
|
Libby Miller,
|
|
Guus Schreiber and
|
|
Michael Uschold.
|
|
</p>
|
|
|
|
<p>
|
|
This document is a product of the Multimedia Annotation on the
|
|
Semantic Web Task Force of the Semantic Web Best Practices and
|
|
Deployment Working Group.
|
|
</p>
|
|
|
|
</body>
|
|
</html>
|