slides 32.2 KB

Raw Blame History Permalink

<?xml version="1.0" encoding="UTF-8"?><!--*- nxml -*-->
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
  "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head profile="http://www.w3.org/2003/g/data-view">
  <title>Semantic Web Data Integration with hCalendar and GRDDL</title>
  <meta name="copyright"
	content="Copyright &#169; 2005 W3C (MIT, ERCIM, Keio)
		 $Revision: 1.33 $" />
  <meta name="Date" content="$Date: 2007/07/10 15:27:52 $" />
  <link rel="transformation"
	href="http://www.w3.org/2002/12/cal/glean-hcal.xsl" />
  <link rel="stylesheet" href="http://www.w3.org/Talks/Tools/Slidy/slidy.css"
	type="text/css"
	media="screen, projection, print" />
  <link rel="stylesheet" href="http://www.w3.org/Talks/Tools/Slidy/w3c-blue.css"
	type="text/css" media="screen, projection, print" />
  <script src="http://www.w3.org/Talks/Tools/Slidy/slidy.js"
	  type="text/javascript">
  </script>

  <style type="text/css">
    .source { text-align: right; font-size: smaller }
    .footnote { font-size: smaller }
    div.figure { text-align: center }
    pre b { color: blue }
    blockquote { border-left: double; padding-left: 1em; font-style: italic; text-align: justify }
  </style>
</head>
<body>

<div class="background">
  <img alt="" id="head-icon"
       src="http://www.w3.org/Talks/Tools/Slidy/icon-blue.png" />
  <object id="head-logo"
	  data="http://www.w3.org/Talks/Tools/Slidy/w3c-logo-blue.svg"
	  type="image/svg+xml" title="W3C logo">
    <a href="http://www.w3.org/"><img
    alt="W3C logo" id="head-logo-fallback"
    src="http://www.w3.org/Talks/Tools/Slidy/w3c-logo-blue.gif" /></a>
  </object>
</div>

<div class="background slanty">
  <img src="http://www.w3.org/Talks/Tools/Slidy/w3c-logo-slanted.jpg"
       alt="slanted W3C logo" />
</div>

<div class="slide cover">
  <img align="right"
       src="http://www.w3.org/Talks/Tools/Slidy/keys.jpg"
       alt="Cover page images (keys)" class="cover" />

<h1>Semantic Web Data Integration with hCalendar and GRDDL</h1>

<address class="vcard">
<a class="url fn n" href="http://www.w3.org/People/Connolly/">
<span class="given-name">Dan</span>
<span class="family-name">Connolly</span></a><br />
</address>

<div class="vevent">
<a class="url summary" href="http://2005.xmlconference.org/">XML
Conference &amp; Exposition 2005 | From Syntax to Semantics (XML
2005)</a><br />
<span class="location">Atlanta, GA, USA</span><br />
<abbr class="dtstart" title="2005-11-16">16 November 2005</abbr>
</div>
</div>


<div class="slide"><h1>Toward Open Data</h1>
<ul>
  <li><blockquote><p>I want my data back.</p>
  <div class="source"><cite>Jon Bosak circa 1997</cite></div>
  </blockquote></li>
  <li><blockquote>
    <p>I've long believed that customers of any application own the
  data they enter into it.</p>
  <div class="source"><cite><a
  href="http://www.veen.com/jeff/archives/000810.html">Jeffrey Veen 2
  November 2005</a></cite></div>
  </blockquote>
</li>
</ul>

<p><em>Is this what Web 2.0 is all about? If so, maybe it's not such a
bad thing.</em></p>

</div>

<div class="slide"><h1>Outline</h1>

<ol>
  <li>A history of (X)HTML and the Web</li>
  <li>Toward data integration, Web style</li>
  <li>Semantic Web * iCalendar = RDF Calendar </li>
  <li>GRDDL: Semantic Web data in XHTML</li>
  <li>hCalendar and microformats</li>
  <li>hCalendar * GRDDL = RDF Calendar</li>
  <li>Microformats + SPARQL</li>
  <li>Microformats + OWL, rules</li>
</ol>

<p><strong>Objective:</strong> making it cost-effective to
record and share knowledge <em>formally</em>, i.e. so that computers can
manipulate it.</p>

</div>

<div class="slide"><h1>Getting into the Web</h1>
<p>What was the tipping point&#8212; the <q>killer app</q>&#8212;<em>for you</em>?</p>

<ul class="incremental">
<li>for Tim Berners-Lee's colleagues at CERN: a cross-platform phone
directory</li>
<li>for me at Convex in 1991: a nifty hypertext system design:
<blockquote>
<pre>
Newsgroups: alt.hypertext
Subject: WorldWideWeb: Summary
Date: 6 Aug 91 16:00:12 GMT

		WorldWideWeb - Executive Summary


The WWW project merges the techniques of information retrieval and hypertext to
make an easy but powerful global information system. ...
</pre>
</blockquote>
</li>
<li>in the early 1990s: NCSA What's New</li>
<li>these days: maps, auctions, wikipedia, webmail, ...</li>
</ul>
</div>

<div class="slide"><h1>Getting into the Web: downhill steps</h1>

<p>To grow, start with <em>some actual value</em> plus lots of
potential value and a downhill path for contributors:</p>

<ul class="incremental">
<li>for a little investment
<p>Note: licensing and legal agreements are <em>not little</em>.</p>
</li>
<li>a big return</li>
<li>due to network effects</li>
<li>Scaling
<ul class="incremental">
  <li>yes, scalability to 10^9 nodes and up is important</li>
  <li>but so is scalability down to families and scout troops</li>
</ul>
</li>
</ul>

<p><q>Architect for participation</q> -- <cite>Tim O'Reilly @ W3C's 10 anniversary</cite></p>
</div>

<div class="slide"><h1>Web Basics: URIs, HTTP, ...</h1>
<p>So what makes all that work? Much of it is a story for another day.</p>
<p>See <cite><a href="http://www.w3.org/TR/2004/REC-webarch-20041215/">Architecture of the World Wide Web, Volume One</a></cite><br />
W3C Recommendation 15 December 2004
</p>
</div>

<div class="slide"><h1>Web Basics: HTML</h1>

<ul>
<li>In the "garage" at CERN:<br />
NextStep RTF object meets SGML</li>
<li>IETF HTML Working Group:<br />
<em>Is <tt>&lt;p></tt> a container or a separator?</em></li>
<li>HTML Validation Service at Halsoft</li>
<li>W3C HTML Working Group:<br />
Netscape trades <tt>&lt;blink></tt> for Microsoft's <tt>&lt;marquee></tt><br />
HTML 3.2 is born</li>
</ul>

</div>

<div class="slide"><h1>Aside: CSS deployment</h1>
<p>...  and the importance of test suites</p>
<ul class="incremental">
  <li>Dec 1996: <a href="http://www.w3.org/TR/REC-CSS1-961217">Original CSS
  Rec</a></li>
  <li>Goodbye <tt>&lt;FONT></tt>, right?</li>
  <li>At least <tt>&lt;FONT></tt> didn't crash my browser</li>
  <li>
    <blockquote>
      <p>The experience of the CSS working group, with the CSS1 Test
      Suite, and the many compliant and interoperable implementations
      that followed, demonstrated fairly clearly that a simple test
      suite is far far better than none at all.</p>
      <div class="source"><cite><a
      href="http://www.w3.org/Style/CSS/Test/testsuitedocumentation.html#something"
      >CSS Test Suite Documentation</a></cite></div>
    </blockquote>
  </li>
</ul>
</div>

<div class="slide"><h1>HTML / SGML * XML = XHTML</h1>
<p>along with CSS, of course</p>
<ul>
  <li>HTML 4.0: a vocabulary we can understand and agree on</li>
  <li>XML 1.0: a syntax we can understand and agree on</li>
  <li>CSS: a style sheet mechanism with graceful degradation</li>
  <li>After ~5 years of single-pixel-gifs and other black magic...</li>
  <li class="incremental">XHTML + CSS arrives<br />
  <cite><a href="http://www.zeldman.com/dwws/">Designing With
  Web Standards</a></cite> by Zeldman, May 2003
  </li>
</ul>
</div>


<div class="slide"><h1>The Personal Information Disaster</h1>

<ul>
<li>For data, we are pre-web:
<dl>
  <dt>Airline sends me email from their database</dt>
    <dd>I copy/paste each of the data into my PDA</dd>
  <dt>Soccer coach distributes a schedule</dt>
    <dd>each players with an online calendar re-keys the data</dd>
</dl>
</li>
<li>
<blockquote>
<p>The bane of my existence is doing things I know the computer could do for me.</p>
<div class="source"><cite><a
href="http://www.nature.com/nature/webmatters/xml/xml.html">The XML Revolution</a></cite>, Nature Web Matters Oct 1998</div>
</blockquote>
</li>
<li>Let's find ways to make it cost-effective record and share
knowledge <em>formally</em>, i.e. so that computers can manipulate
it.</li>
</ul>

</div>

<div class="slide"><h1>RDF Calendar for travel itineraries</h1>
<img align="right"  alt="cal screen shot"
     src="http://www.w3.org/2000/10/swap/pim/calIntShot.png" />
<ul>
  <li>travel admin mails text itinerary
  <ul>
    <li>dumped from database, probably</li>
  </ul>
  </li>
  <li>nasty perl script scrapes RDF statements in travel terms</li>
  <li>logical rules convert to iCalendar vocabulary
  <ul><li>mixing in airport timezone info</li></ul>
  </li>
  <li>toIcal.py converts RDF calendar to iCalendar .ics syntax</li>
</ul>
<p><em>But why not just spit out .ics from the nasty perl script?</em></p>
</div>

<div class="slide"><h1>Travel from place to place</h1>

<img align="right" alt="map of itinerary"
  src="../../../../2003/04dc-mia/itin-mia.png" />
<ul>
<li>for details, see <a href="../../../../2000/10/swap/pim/travel">Integration Example: travel tools</a> in <cite><a href="../../../../2000/10/swap/doc/">Semantic Web Tutorial Using N3</a></cite> given at WWW2003 in Budapest</li>
</ul>

<p class="footnote">iCalendar does have location/geo fields, but only
one per event. Flights have a departure and an arrival.</p>

</div>

<div class="slide"><h1>TODO list views of email archives</h1>

<img align="right" src="http://lists.w3.org/Archives/Public/www-archive/2004Feb/att-0080/n3bugstodo-grab.png" alt="iCal screenshot" />
<ul>
<li>lists.w3.org gives message headers in RDF</li>
<li>convention: <tt>[closed]</tt> marks closed threads in public-cwm-bugs</li>
<li>some rules relate these conventions to RDF/iCalendar</li>
<li>toIcal.py converts to .ics format</li>
<li>developers subscribe to TODO list</li>
</ul>

<p class="footnote">details: <a
href="http://lists.w3.org/Archives/Public/www-rdf-calendar/2004Jan/0011.html">bug
status in .ics</a> of Jan 2004</p>
</div>

<div class="slide"><h1>Semantic Web Basics</h1>
<ul>
<li>What is the Semantic Web?
<ul>
 <li>Data integration across application, organizational boundaries</li>
</ul>
</li>
<li>How does it work?
<ol>
<li>Apply power of URIs to concepts of relational data
<ul>
<li>Don't say "colour" say &lt;http://example.com/2002/std6#col></li>
</ul>
</li>
<li>Model real things, not just documents or database tables</li>
</ol>
</li>
</ul>
</div>

<div class="slide"><h1>The element of the Semantic Web</h1>

<p><img alt="arrow tail, body and head are l are subject, property and value." src="../../../../DesignIssues/diagrams/spv-arrow.png" /></p>

<ul>
  <li>The Resource Description Framework (RDF)
  <ul>
    <li>abstract syntax, formal semantics</li>
    <li>standard encoding in XML</li>
    <li>emerging programmer-friendly short-hand encoding: <a href="http://www.w3.org/2000/10/swap/Primer.html">N3</a>/<a href="http://www.dajobe.org/2004/01/turtle/">turtle</a>
    </li>
  </ul>
  </li>

</ul>
    <pre>
      @prefix foaf: &lt;http://xmlns.com/foaf/0.1/>.
      &lt;people.rdf#dan> foaf:name "Dan Connolly".
      &lt;people.rdf#dan> foaf:interest &lt;http://www.w3.org/XML/>.
    </pre>
<p>Note the relationship to HTML links, especially with the
re-discovery of the <tt>rel</tt> attribute.</p>

</div>

<div class="slide"><h1>Semantic web includes tables, trees...</h1>

<p><img alt="Arrows can make a table, an arrow from each row to each value"
 src="../../../../DesignIssues/diagrams/arrow-table.png" /></p>

<p><img alt="Arrows can make a table, an arrow from each row to each value"
 src="../../../../DesignIssues/diagrams/tree.png" /></p>

</div>

<div class="slide"><h1>... and tangly messes</h1>

<p><img alt="Arrows can make a table, an arrow from each row to each value"
src="../../../../DesignIssues/diagrams/tree-and-table2.png" /></p>

</div>

<div class="slide"><h1>RDFS and OWL</h1>

<p>RDF Schema (RDFS) and the Web Ontology Language (OWL) correspond to
UML notions such as subclass, domain, range, cardinality, ...</p>

<img align="right"
     src="http://www.w3.org/2000/10/swap/pim/travelFig.png"
     alt="travel concepts schema" />

<ul>
<li style="color: red;">cyc terms  in red</li>
<li style="color: purple;">DAML airport ontology terms in purple</li>
<li style="color: green;">custom travelTerms in green.</li>
<li style="color: blue;">RDF standard terms in blue</li>
<li style="color: orange;">XML Schema terms in orange</li>
</ul>

</div>

<div class="slide"><h1>RDF Calendar in a nutshell</h1>

<ul>

<li>In iCalendar terms (<a href="../rfc2445">RFC2445</a>), an event is
a <a href="../rfc2445#sec4.6">component</a>
with various <a
href="../rfc2445#sec4.5">properties</a>:

<pre>BEGIN:VEVENT
<span xml:lang="en" lang="en">UID:20020630T230445Z-3895-69-1-7@jammer</span>
<span xml:lang="en" lang="en">DTSTART;VALUE=DATE:20020703</span>
<span xml:lang="en" lang="en">DTEND;VALUE=DATE:20020706</span>
<span xml:lang="en" lang="en">SUMMARY:Scooby Conference</span>
<span xml:lang="en" lang="en">LOCATION:San Francisco</span>
<span xml:lang="en" lang="en">END:VEVENT</span></pre>
</li>
<li>and RDF/XML has analogous class and property constructs:
<pre>      &lt;Vevent&gt;
        &lt;uid&gt;20020630T230445Z-3895-69-1-7@jammer&lt;/uid&gt;
        &lt;dtstart&gt;2002-07-03&lt;/dtstart&gt;
        &lt;dtend&gt;2002-07-06&lt;/date&gt;
        &lt;summary&gt;Scooby Conference&lt;/summary&gt;
        &lt;location&gt;San Francisco&lt;/location&gt;
      &lt;/Vevent&gt;</pre>
</li>
</ul>

<p>For details, see <cite><a
href="http://www.w3.org/TR/rdfcal/">RDF Calendar - an application of
the Resource Description Framework to iCalendar Data</a></cite>,
Connolly and Miller September 2005</p>
</div>

<div class="slide"><h1>Round-trip testing</h1>

<p>Comparing .ics files is tricky, so...</p>

<ul>
  <li>to test .ics -> RDF:
  <ol>
   <li>convert <tt>$testcase.ics</tt> to temporary <tt>$testcase-actual.rdf</tt></li>
   <li>Use RDF graph comparison to check actual results vs. expected results</li>
  </ol>
  </li>
  <li>to test RDF -> .ics<br />
  w.r.t a known-good .ics -> RDF tool:
  <ol>
    <li>convert <tt>$testcase.rdf</tt> to <tt>$testcase-temp.ics</tt></li>
    <li>Use the known-good tool get <tt>$testcase-actual.rdf</tt></li>
    <li>Use RDF graph comparison to check <tt>$test-case-actual.rdf</tt> against expected results</li>
  </ol>
  </li>
</ul>
</div>

<div class="slide"><h1>On RDF/XML Syntax</h1>
<ul>
 <li>Too constrained for some
  <ul>
    <li>Tries to look like Book/author/title metadata but has subtle
    constraints with striping, rdf:about,rdf:resource, etc.</li>
  </ul>
 </li>
 <li>Not constrained enough for others
  <ul>
    <li>doesn't work with DTDs nor W3C XML Schemas; works awkwardly
    with XPath and XSLT</li>
  </ul>
 </li>
</ul>
<p>At least the issues in the 1998 spec have all been resolved,
complete with test cases. There are plenty of interoperable
parsers. And it works great with Relax-NG and nxml-mode :)</p>
</div>

<div class="slide"><h1>Observation: lots of structured data in XHTML
dialects</h1>

<ul>
<li>bibliographies</li>
<li>schedules</li>
<li>issues lists</li>
<li>news items</li>
<li>blog rolls</li>
</ul>
</div>

<div class="slide"><h1>Data in Documents</h1>

<blockquote>
<p>I believe that one of the best ways to transition into RDF,
if not a long-term deployment strategy for RDF, is to manage the
information in human-consumable form (XHTML) annotated with just
enough info to extract the RDF statements that the human info
is intended to convey. In other words: using a relational
database or some sort of native RDF data store, and spitting
out HTML dynamically, is a lot of infrastructure to operate
and probably not worth it for lots of interesting cases.</p>

<p>We all know that we have to produce a human-readable version
of the thing... why not use that as the primary source?</p>
<div class="source">
<cite><a href="http://lists.w3.org/Archives/Public/www-rdf-interest/2000Mar/0103.html">XSLT for screen-scraping RDF out of real-world data</a><br />Dan Connolly to www-rdf-interest March 2000</cite>
</div>
</blockquote>
</div>

<div class="slide"><h1>Case Study: News Syndication at W3C</h1>
<ul>
 <li>W3C Communications Team knows HTML well; picked up XHTML quickly</li>
 <li>RSS emerged as a syndication mechanism</li>
 <li>W3C's news content management system is just CVS; no database</li>
 <li>But by adding just a <tt>class</tt> attribute here and a
 <tt>rel</tt> attribute there, we get enough structure to transform it
 to RSS using XSLT</li>
</ul>

<p><cite><a href="../../../../2000/08/w3c-synd/">Site Summaries
in XHTML</a></cite> is a cost-effective way to formalize our news
metadata.</p>
</div>

<div class="slide"><h1>GRDDL Semantics: explicit, grounded in the Web</h1>
<ul>
  <li>Screen scraping: at your own risk</li>
  <li>GRDDL: author explicitly agrees to data extraction algorithm</li>

  <li><tt>rel="transformation"</tt> is grounded in URI space: <br />
  <tt>http://www.w3.org/2003/g/data-view#transformation</tt>
  </li>

  <li>output of transformation is expressed in RDF, using URIs for
  property names</li>

</ul>
  <p>A person (or a machine) can "follow your nose" from the document
  to the transformation algorithm, to the data, to the definitions of
  the terms used in the data.</p>
</div>


<div class="slide"><h1>GRDDL Syntax: author's choice</h1>

<ul>
  <li>works with DTD-happy XHTML<br />
  via a link with <tt>rel="transformation"</tt>,
  qualified by the GRDDL profile</li>
  <li>works with XML in general<br />
  via <tt>grdd:transformation</tt> attribute on the root element</li>
  <li>the link refers to a transformation, typically XSLT, that maps
  to RDF/XML</li>
  <li>the link may be indirect via a profile or namespace document</li>
</ul>

</div>


<div class="slide"><h1>GRDDL drawback: turing completeness</h1>

<ul>
  <li>XSLT is a big hammer
  <ul>
    <li>much less so than in 2000</li>
  </ul>
  </li>
  <li>Trust issues: consumer runs XSLT from <em>who knows where</em>
  <ul>
    <li>well-known transformations should cover 80%
    <ul>
      <li>short-cut local implementation</li>
      <li>known-good cached copies</li>
    </ul>
    </li>
  </ul>
  </li>
</ul>
</div>


<div class="slide"><h1>GRDDL Details</h1>

<ul>
  <li>In addition to structured XHTML, add the GRDDL profile and
  a typed link:

<pre>&lt;html xmlns="http://www.w3.org/1999/xhtml"&gt;
&lt;head <b>profile="http://www.w3.org/2003/g/data-view"</b>&gt;
  <b>&lt;link rel="transformation"
     href="http://www.w3.org/2000/08/w3c-synd/home2rss.xsl" /&gt;</b>
  ...
&lt;/head>
&lt;body>
  ...
&lt;div id="x200501110b" class="item">
&lt;h3>&lt;img alt="" width="17" height="11" src="/Icons/right" />Policy for
Authorized W3C Translations Announced&lt;/h3>

&lt;p>&lt;span class="date">2005-11-10:&lt;/span> W3C is pleased to announce...</pre>
  </li>

  <li>The <tt>home2rss</tt> transformation turns that into RDF/XML:
<pre>...
&lt;item rdf:about="http://www.w3.org/News/2005#item158">
&lt;title>Policy for Authorized W3C Translations Announced&lt;/title>
&lt;description>2005-11-10: W3C is pleased to announce ...</pre>
  </li>
</ul>
</div>

<div class="slide"><h1>GRDDL: multiple dialects allowed</h1>

<div class="figure">
<img src="../../../../2004/01/rdxh/figMultiTxform.png"
 alt="one document, multiple transformations"/>
</div>
</div>

<div class="slide"><h1>GRDDL: one profile for lots of documents</h1>

<div class="figure">
<img src="../../../../2004/01/rdxh/figGleanProfile.png"
 alt="transformation via profile"/>
</div>

<p class="footnote">For details, see <cite><a
href="http://www.w3.org/TR/grddl/">Gleaning Resource Descriptions from
Dialects of Languages (GRDDL)</a></cite>, Haza&#235;l-Massieux and
Connolly May 2005</p>
</div>


<div class="slide"><h1>GRDDL for any XML, not just XHTML</h1>

<pre>
&lt;java version="1.5.0_04" class="java.beans.XMLDecoder"
   <b>xmlns:grddl="http://www.w3.org/2003/g/data-view#"
   grddl:transformation="<a href="http://www.w3.org/2001/tag/2005/09/grokVioletUML.xsl">grokVioletUML.xsl</a>"></b>
 &lt;object class="com.horstmann.violet.ClassDiagramGraph">
  &lt;void method="addNode">
   ...
</pre>

<div class="figure">
<img src="http://www.w3.org/2001/tag/2005/09/20-class-uml.png"
     alt="UML diagram with OWL formalization" />
</div>

</div>

<div class="slide">
<h1>Origins of hCalendar and Microformats</h1>

<p><em>Borrowing from <a
href="http://tantek.com/presentations/2005/09/microformats-evolution/">Microformats:
Evolving the Web</a> by Tantek Çelik Sep 2005</em></p>

<ul class="incremental">
<li>Very common "structures" on weblogs</li>
<li>Create explicit structures
<ul><li>in order to easily publish, index, aggregate</li></ul>
</li>
<li>Minimize impact on authors (and developers)</li>
<li>Avoid duplicating all content
 <ul><li>Avoid requiring file uploads</li></ul>
</li>
</ul>

</div>

<div class="slide">
<h1>Microformats Principles</h1>
<ul class="incremental">
<li>solve a specific problem</li>
<li>simple as possible
<ul class="incremental">
<li>evolutionary improvements</li></ul>
</li>
<li>humans first, machines second

<ul class="incremental">
<li>presentable <em>and</em> parsable</li>
<li>adapt to current behaviors</li>
</ul>
</li>
<li>reuse from widely adopted standards
<ul class="incremental">
<li>semantic (X)HTML, schemas from interoperable RFCs</li>
</ul>
</li>
<li>modularity / embeddability</li>

<li>decentralized development, content, services
<ul><li>explicitly encourage "spirit of the Web"</li></ul>
</li>
</ul>
</div>

<div class="slide">
<h1>XHTML * iCalendar = hCalendar</h1>
<ul>
<li>Map iCalendar constructs 1:1 into XHTML
<ul>
<li><a href="http://microformats.org/wiki/hcalendar">hCalendar</a> specification, issues, examples... wiki</li>
</ul>
</li>

<li>example
<pre><code>
&lt;ol <span class="added">class="vcalendar"</span>&gt;
 &lt;li <span class="added">class="vevent"</span>&gt;
  &lt;a href="http://tantek.com/presentations/..."
     <span class="added">class="url"</span>&gt;

   &lt;span <span class="added">class="summary"</span>&gt;Microformats: Evolving the Web&lt;/span
   &lt;abbr <span class="added">class="dtstart" title="20050930T1530+1000"</span>&gt;
    September 30th, 2005
   &lt;/abbr&gt;
  &lt;/a&gt;
 &lt;/li&gt;

&lt;/ol&gt;
</code></pre>
</li></ul>
</div>

<div class="slide">
<h1 style="text-transform:none">more hCalendar</h1>
<ul>
<li><a href="http://theryanking.com/microformats/hcalendar-creator.html">hCalendar creator</a></li>
<li>This presentation</li>
<li><a href="http://concerts.shrub.ca/">Sunnyvale House Concerts site</a></li>
<li><a href="http://evdb.com/">EVDB</a> and <a href="http://upcoming.org/">Upcoming.org</a></li>

<li><a href="http://favelets.com/#microformats">Favelets</a> for hCalendars</li>
<li><a href="http://tantek.com/presentations/2005/09/microformats-evolution/we05-program.html">Web Essentials 05 Program</a></li>
</ul>
</div>


<div class="slide"><h1>hCalendar + GRDDL = RDF Calendar</h1>

<ul>
  <li><a href="e1.html">hCalendar</a> + <a href="../glean-hcal.xsl">glean-hcal.xsl</a> GRDDL transformation:<br />
  <iframe width="80%" height="80" src="e1.html">
    <a href="e1.html">event 1</a>
  </iframe>
  </li>
  <li>RDF:
<pre>
@prefix : &lt;http://www.w3.org/2002/12/cal/icaltzd#> .
@prefix XML: &lt;http://www.w3.org/2001/XMLSchema#> .

[       a :Vevent;
           :attendee  [ :cn "Hoopy Frood";
             :calAddress &lt;mailto:frood@example> ];
           :dtend "2005-10-07"^^XML:date;
           :dtstart "2005-10-05"^^XML:date;
           :geo  (37.0625 -95.677068 );
           :location "Argent Hotel, San Francisco, CA";
           :summary "Web 2.0 Conference";
           :url &lt;http://www.web2con.com/> ].
</pre>
  </li>
</ul>
</div>

<div class="slide"><h1>Upcoming events on blogs</h1>

<p>From <a href="http://tantek.com/log/">Tantek's Thoughts</a>:</p>
<blockquote>
<ul>

<li class="vevent">
 <a class="url" href="http://web2con.com/">
 <abbr class="dtstart" title="20051005">
  10/5</abbr>-<abbr class="dtend" title="20051008">7</abbr> <span class="summary">

  Web 2.0 </span>
- at
 <span class="location">
  The Argent, San Francisco </span>
 </a>
</li>

<li class="vevent">
 <abbr class="dtstart" title="20051017">
  10/17</abbr>-<abbr class="dtend" title="20051020">19</abbr>

  <span class="summary">W3C CSS Working Group f2f</span>
  at <span class="location">San Francisco</span>
</li>
</ul>
</blockquote>

<p>Mix with <a href="../glean-hcal.xsl">glean-hcal.xsl</a> and we get:</p>
<pre style="text-size: smaller">
&lt;r:RDF xmlns:r="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:c="http://www
.w3.org/2002/12/cal/icaltzd#">
...
      &lt;c:Vevent>
        &lt;c:summary xml:lang="en-us">Web 2.0 &lt;/c:summary>
        &lt;c:dtstart r:datatype="http://www.w3.org/2001/XMLSchema#date">2005-10-05&lt;/c:dtstart>
        &lt;c:dtend r:datatype="http://www.w3.org/2001/XMLSchema#date">2005-10-08&lt;/c:dtend>
        &lt;c:url r:resource="http://web2con.com/"/>
        &lt;c:location xml:lang="en-us">The Argent, San Francisco &lt;/c:location>
      &lt;/c:Vevent>
</pre>

</div>

<div class="slide"><h1>SQL * URIs = SPARQL</h1>

<p>Use GRDDL to aggregate data from friends etc, then...</p>
<img align="right"
     src="http://www.w3.org/DesignIssues/diagrams/spv-table.png"
     alt="table subject/property/value" />

<pre>
  PREFIX foaf: &lt;http://xmlns.com/foaf/0.1/>
  PREFIX c: &lt;http://www.w3.org/2002/12/cal/icaltzd#>
  SELECT ?name, ?summary, ?when
   FROM &lt;myFriendsBlogsData>
   WHERE { ?somebody foaf:name ?name; foaf:mbox ?mbox.
           ?event c:summary ?summary;
                  c:dtstart ?ymd;
                  c:attendee [ c:calAddress ?mbox ]
         }.
</pre>

<table border="1">
<tr><th>?name</th><th>?summary</th><th>?when</th></tr>
<tr><td>Tantek Çelik</td><td>Web 2.0</td><td>2005-10-05</td></tr>
<tr><td>Norm Walsh</td><td>XML 2005</td><td>2005-11-13</td></tr>
<tr><td>Dan Connolly</td><td>W3C tech plenary</td><td>2006-02-27</td></tr>
</table>

<p>See <cite><a href="http://www.w3.org/TR/rdf-sparql-query/">SPARQL
Query Language for RDF</a></cite> W3C Working Draft 21 July 2005, plus
<a href="http://norman.walsh.name/2005/itinerary/">Norm's travel this
year</a></p>
</div>


<div class="slide"><h1>Fill in blanks with rules</h1>

<p>Mix with some logical rules:</p>
<ul>
<li>Infer the place implicit in the iCalendar geo and location properties.
<pre>
{  ?E c:geo (?LAT ?LONG) }
 => { ?E k:eventOccursAt [ geo:lat ?LAT; geo:long ?LONG] }.

{  ?E c:location ?PLACENAME; k:eventOccursAt ?WHERE
} => {
  ?WHERE k:nameOfAgent ?PLACENAME;
    s:label ?PLACENAME;
    k:inRegion [ a k:City ]. # Assume each event is in some city
}.
</pre>
</li>
<li>Cities containing points far from each other are distinct
<pre>
{
  ?C1 a k:City; is k:inRegion of [geo:lat ?X1; geo:long ?Y1 ].
  ?C2 a k:City; is k:inRegion of [geo:lat ?X2; geo:long ?Y2 ].

  # DSQ = (X1-X2)^2 + (Y1-Y2)^2
  (((?X2 ?X1).m:difference 2).m:exponentiation
   ((?Y2 ?Y1).m:difference 2).m:exponentiation) m:sum ?DSQ.
  ?DSQ m:greaterThan 0.2.
} => { ?C1 owl:differentFrom ?C2 }.
</pre>
</li>
</ul>
<p class="footnote">Full details:
<a href="calbg.n3">calendar background rules</a>
</p>
</div>

<div class="slide"><h1>A model behind hCalendar</h1>

<div class="figure">
<object width="90%" height="400" data="e1c.svg">
 <img src="e1c.png" alt="person, place, dates" />
</object>
</div>

<p>Can we do this event too?</p>

    <iframe width="80%" height="80" src="e2.html">
      <a href="e2.html">event 2</a>
    </iframe>
</div>

<div class="slide"><h1>RDF merges trivially</h1>

<img style="float:right" width="50%" class="incremental"
  src="http://www.w3.org/2000/Talks/0906-xmlweb-tbl/arcs-3.gif"
  alt="stack" />

<img src="http://www.w3.org/2000/Talks/0906-xmlweb-tbl/arcs-1.gif"
     alt="stack" />

</div>

<div class="slide"><h1>Partial Understanding</h1>

<p>RDF statements* are independent. RDF semantics are <dfn><a href=
"http://en.wikipedia.org/wiki/Monotonicity_of_entailment"
>monotonic</a></dfn>.</p>

<table>
  <tr><th></th><th>RDF</th><th>XML</th></tr>
  <tr>
    <th>Premise</th>
    <td>
<pre>
&lt;Book rdf:ID="book1">
 &lt;dc:title>The Grapes of Wrath&lt;/title>
 &lt;dc:creator>Steinbeck&lt;/author>
&lt;/Book>
</pre>
    </td>
    <td>
<pre>
&lt;xsd:simpleType name="myInteger">
  &lt;xsd:restriction base="xsd:integer">
    &lt;xsd:minInclusive value="10000"/>
    &lt;xsd:maxInclusive value="99999"/>
  &lt;/xsd:restriction>
&lt;/xsd:simpleType>
</pre>
    </td>

  </tr>
  <tr><th>Conclusion</th>
  <td>
<pre>
&lt;Book rdf:ID="book1">
 &lt;dc:title>The Grapes of Wrath&lt;/title>
&lt;/Book>
</pre>
  </td>
    <td>
<pre>
&lt;!-- no, this does not follow -->
&lt;xsd:simpleType name="myInteger">
  &lt;xsd:restriction base="xsd:integer">
    &lt;xsd:maxInclusive value="99999"/>
  &lt;/xsd:restriction>
&lt;/xsd:simpleType>
</pre>
    </td>
  </tr>
</table>

<p class="footnote">*RDF/XML does have a
<tt>rdf:parseType="Collection"</tt> syntax, which
expands to a lisp style binary tree in the abstract syntax. This
erasure property works not on XML elements, but on RDF statements.
</p>
</div>

<div class="slide"><h1>Merging hCalendar data</h1>

<div class="figure">
  <object width="90%" height="500" data="mash.svg">
    <img src="mash.png" alt="two events, one person, several days and places" />
  </object>
</div>
</div>

<div class="slide"><h1>OWL identity reasoning</h1>

<table>
  <tr>
    <th>Premise</th>
    <td>
<pre>
# one-to-many
foaf:mbox a owl:InverseFunctionalProperty.

:dan foaf:mbox &lt;mailto:connolly@w3.org>.
:dan foaf:name "Dan Connolly".

:daniel foaf:mbox &lt;mailto:connolly@w3.org>.
:daniel foaf:name "Daniel W. Connolly".
</pre>
    </td>
    <td>
      <object  data="owl_smush1.svg">
	<img style="float:right" alt="" src="owl_smush1.png" />
      </object>
    </td>
  </tr>
  <tr><th>Conclusion:</th>
  <td>
<pre>
:daniel owl:sameAs :dan.
:daniel foaf:name "Dan Connolly".
:daniel foaf:name "Daniel W. Connolly".
</pre>
  </td>
  <td>
     <object  data="owl_smush2.svg">
      <img style="float:right" alt="" src="owl_smush2.png" />
    </object>
  </td>
  </tr>
</table>

</div>

<div class="slide"><h1>Consistency checking</h1>


<ul>
<li>Modelling assumption: on a given day, a person can only be in one city
<pre>
PersonDay a owl:Class; s:subClassOf k:Entity;
 s:subClassOf [ owl:onProperty _city; owl:maxCardinality 1 ].
</pre>
</li>

<li>Then run <a
href="http://www.mindswap.org/2003/pellet/index.shtml">pellet</a>, an
OWL consistency checker:
<pre>
java -jar pellet.jar -inputFile ,mash.rdf -unsat
Input file: ,mash.rdf
OWL Species: Full
DL Expressivity: ALCHIF(D)
Consistent: No
Reason: There is an anonymous individual X, identified by
  this path (file:...cal/mash/calbg#<b>oct6</b>
      http://www.cyc.com/2004/06/04/cyc#temporallyIntersects X),
  which has <b>more than 1</b> values for property file:...cal/mash/calbg#<b>_city</b>
  violating the cardinality restriction
Time: 8784 ms (Loading: 6485 Preprocessing: 0
   Species Validation: 2202 Consistency: 96 )
</pre>
</li>
<li>In other words: <em>You can't be two places at once!</em></li>
</ul>

</div>

<div class="slide"><h1>Review</h1>

<ul>
 <li>hCalendar is not too much to ask of authors</li>
 <li>GRDDL makes it authorized Semantic Web data</li>
 <li>Semantic Web data:
 <ul class="incremental">
   <li>compares straightforwardly</li>
   <li>merges easily</li>
   <li>can be queried with SPARQL</li>
   <li>can be checked for consistency with OWL reasoners</li>
   <li>mixes with emerging rule languages</li>
 </ul>
 </li>
</ul>
</div>

<div class="slide"><h1>Acknowledgements, Colophon</h1>

<p>These slides: <tt><a
href="http://www.w3.org/2002/12/cal/mash/slides">http://www.w3.org/2002/12/cal/mash/slides</a></tt></p>

<ul>
  <li>Dog-food: view source for...
  <ul>
    <li>hCalendar</li>
    <li>hCard</li>
    <li>GRDDL</li>
  </ul>
  </li>
  <li>written using nxml-mode</li>
  <li>CSS/javascript presentation magic is <a href="http://www.w3.org/Talks/Tools/Slidy/help.html">Slidey</a></li>

<li>References

<ul>
<li><cite><a href="../../../../2000/10/swap/doc/">Semantic Web Tutorial Using N3</a></cite> given at WWW2003 in Budapest</li>
<li><cite><a href="http://www.dajobe.org/2004/01/turtle/">Turtle - Terse RDF Triple Language</a></cite> Dave Beckett, in progress 2004-2005</li>
<li><a href="http://esw.w3.org/topic/RdfCalendarPresentation">RDF calendar presentation</a>, Semantic Web Interest Group Meeting March 2004</li>
<li>See <cite><a href="http://www.w3.org/TR/2004/REC-webarch-20041215/">Architecture of the World Wide Web, Volume One</a></cite><br />
W3C Recommendation 15 December 2004</li>
<li><cite><a
href="http://www.w3.org/TR/grddl/">Gleaning Resource Descriptions from
Dialects of Languages (GRDDL)</a></cite>, Haza&#235;l-Massieux and
Connolly May 2005</li>
<li><cite><a href="http://www.w3.org/TR/rdf-sparql-query/">SPARQL Query Language for RDF</a></cite>
W3C Working Draft 21 July 2005</li>
<li><cite><a
href="http://www.w3.org/TR/rdfcal/">RDF Calendar - an application of
the Resource Description Framework to iCalendar Data</a></cite>,
Connolly and Miller September 2005</li>
</ul>
</li>
</ul>
</div>
  </body>
</html>