CINXE.COM

Feedback loops and formal systems

<?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 http://purl.org/NET/erdf/profile"> <title>Feedback loops and formal systems</title> <meta name="copyright" content="Copyright &#169; 2005 W3C (MIT, ERCIM, Keio) $Revision: 1.6 $" /> <meta name="Date" content="$Date: 2006/03/01 13:51:00 $" /> <link rel="transformation" href="https://www.w3.org/2002/12/cal/glean-hcal.xsl" /> <link rel="transformation" href="https://norman.walsh.name/2005/12/12/examples/hcard2rdf.xsl" /> <link rel="foaf-maker" href="#danc" /> <link rel="foaf-topic" href="#tp2006" /> <link rel="schema.foaf" href="https://xmlns.com/foaf/0.1/" /> <link rel="stylesheet" href="https://www.w3.org/Talks/Tools/Slidy/slidy.css" type="text/css" media="screen, projection, print" /> <link rel="stylesheet" href="https://www.w3.org/Talks/Tools/Slidy/w3c-blue.css" type="text/css" media="screen, projection, print" /> <!-- embeded RDF txform seems to need this --> <base href="https://www.w3.org/2006/Talks/0301-dc-afm/" /> <script src="https://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="https://www.w3.org/Talks/Tools/Slidy/icon-blue.png" /> <object id="head-logo" data="https://www.w3.org/Talks/Tools/Slidy/w3c-logo-blue.svg" type="image/svg+xml" title="W3C logo"> <a href="https://www.w3.org/"><img alt="W3C logo" id="head-logo-fallback" src="https://www.w3.org/Talks/Tools/Slidy/w3c-logo-blue.gif" /></a> </object> </div> <div class="background slanty"> <img src="https://www.w3.org/Talks/Tools/Slidy/w3c-logo-slanted.jpg" alt="slanted W3C logo" /> </div> <div class="slide cover"> <img align="right" src="https://www.w3.org/Talks/Tools/Slidy/keys.jpg" alt="Cover page images (keys)" class="cover" /> <h1>Feedback loops and formal systems</h1> <address class="vcard" id="danc"> <a class="url fn n foaf-name" rel="foaf-homepage" href="https://www.w3.org/People/Connolly/"><span class="given-name">Dan</span > <span class="family-name">Connolly</span></a><br /> </address> <div class="vevent" id="tp2006"> <a class="url summary foaf-homepage" href="https://www.w3.org/2005/12/allgroupoverview.html">W3C Technical Plenary</a><br /> <span class="location">Mandelieu, France</span><br /> <abbr class="dtstart" title="2006-03-01">1 March 2006</abbr> </div> </div> <div class="slide"><h1>Classical vs Chaotic</h1> <ul> <li>Research Interest: <q>... investigating the value of formal descriptions of chaotic systems like the Web, especially in the consensus-building process.</q></li> <li>Predicting change in the Web is more like predicting the weather or financial markets than predicting the path of a bullet.</li> <li>But I want my bank to use classical reasoning, not heuristics, when computing my balance</li> </ul> </div> <div class="slide"><h1>Statistical Approaches are Great</h1> <ul> <li>PageRank</li> <li>tagging a la flickr, del.icio.us</li> <li>...<q>wisdom of crowds</q></li> </ul> <p>I'm learning that stuff, slowly, but symbolic math is my strength.</p> </div> <div class="slide"><h1>HTML 1.x</h1> <ul> <li>1990: Convex/HP SGML work was an excuse to work on HTML: <ol> <li>draft a DTD</li> <li>try it on some test cases</li> <li>discover an issue; repeat</li> </ol> </li> </ul> </div> <div class="slide"><h1>HTML 2.0</h1> <ul> <li>1994: adding HTML to products at HalSoft</li> <li>Asked others to try <tt>sgmls</tt> parser; not many takers</li> <li>Installed <tt>sgmls</tt> as a CGI service; HTML validator is born</li> <li>feedback loop works over the Web</li> </ul> </div> <div class="slide"><h1>Are DTDs cost-effective?</h1> <ul> <li>To SGML or not to SGML? <div>what if we had just used lex/yacc?</div> </li> <li>DTD syntax used to be obscure; less so now</li> <li>What's wrong with a <tt>&lt;form></tt> that spans 3 <tt>&lt;tr></tt>s? <ul> <li>nothing, really, but it's hard to say in a DTD</li> </ul> </li> <li>And now XML Schema languages are facilitating a lot of feedback loops...</li> </ul> </div> <div class="slide"><h1>XSD Schemas put structures in the Web</h1> <ul> <li>Not just <tt>&lt;purchase-order></tt>, but</li> <li> <tt>&lt;ubl:purchase-order></tt> <br />where... <ul> <li>ubl: is grounded in the web at<br /> <tt>http://docs.oasis-open.org/ubl/cd-UBL-1.0/xsd/common/...</tt> </li> </ul> </li> </ul> <p>Now I can link to it, not just informally from hypertext, but formally from other XSD documents...</p> </div> <div class="slide"><h1>Formal Mix and Match</h1> <ul> <li>Compose <tt>&lt;ubl:purchase-order></tt> with <tt>&lt;pharma:drug-interactions></tt> <br />where... <ul> <li>pharma: is (hypothetically) grounded in the web at<br /> <tt>http://pharma.example/drugml/...</tt></li> </ul> </li> <li>When does it work? The <tt>pharma:</tt> folks clearly know about XSD; do they have to know about <tt>ubl:</tt>?</li> </ul> </div> <div class="slide"><h1>Circles and Arrows on the Whiteboard</h1> <p>Tag had a circles-and-arrows diagram of Language Evolution (extensibility and versioning):</p> <div class="figure"> <img src="https://www.w3.org/2001/tag/2005/09/20-whiteoard/20050922-093038.jpg-tmb.jpg" alt="Dan at whiteboard" /> </div> </div> <div class="slide"><h1>Cirlces and Arrows in Drawing Tools</h1> <img src="https://www.w3.org/2001/tag/2005/09/22-diagram1.png" alt="diagram" /> <p>TimBL captured it in OmniGraffle. <span class="footnote">See also: <a href="https://www.w3.org/2001/tag/2005/09/22-diagram1.graffle">xml/graffle source</a></span>.</p> </div> <div class="slide"><h1>UML Diagrams</h1> <img src="https://www.w3.org/2001/tag/2006/ext-vers/ext-vers-uml.png" alt="uml diagram" /> </div> <div class="slide"><h1>UML to OWL</h1> <img src="tag-uml-tabu.png" alt="tag vocab as OWL in tabulator"/> <p>Extract RDF from the UML diagram...</p> <pre> <a href="https://www.w3.org/2001/tag/2006/ext-vers/">ext-vers</a>$ make test-agent-pf.n3 xsltproc ... --output ext-vers-uml.rdf grokVioletUML.xsl ext-vers-uml.violet </pre> <p class="footnote">See <cite><a href="http://dig.csail.mit.edu/breadcrumbs/node/87">Using RDF and OWL to model language evolution</a></cite> for details.</p> </div> <div class="slide"><h1>Colophon, References</h1> <p>These slides: <tt><a href="https://www.w3.org/2006/Talks/0301-dc-afm/">http://www.w3.org/2006/Talks/0301-dc-afm/</a></tt></p> <p></p> <ul> <li>written using nxml-mode; CSS/javascript presentation magic is <a href="https://www.w3.org/Talks/Tools/Slidy/help.html">Slidey</a></li> <li>view source for hCalendar, hCard, GRDDL...</li> </ul> <h2>References</h2> <ul> <li><cite><a href="http://dig.csail.mit.edu/breadcrumbs/node/87">Using RDF and OWL to model language evolution</a></cite> 2006-02-15</li> </ul> <h2>See Also</h2> <ul> <li><a href="http://dig.csail.mit.edu/breadcrumbs/node/85">bnf2turtle -- write a RDF/turtle version of an EBNF grammar</a>, 2006-02-10</li> <li><a href="https://www.w3.org/2000/Talks/www9-larch/all.htm">Specifying Web Architecture with Larch</a> given at WWW9 in May 2000 in Amsterdam</li> </ul> </div> </body> </html>

Pages: 1 2 3 4 5 6 7 8 9 10