Conference

I was pleased to find that AAAI '06 in Boston a couple weeks ago had a spectrum of people I know and don't know and work that's near and far from my own. The talk about the DARPA grand challenge was inspiring.

But closer to my work, I ran into Jeff Heflin, who I worked with on DAML and especially the OWL requirements document. Amid too many papers about ontologies for the sake of ontologies and threads like Is there real world RDF-S/OWL instance data?, his Investigation into the Feasibility of the Semantic Web is a breath of fresh air. The introduction sets out their approach this way:

Our approach is to use axioms of OWL, the de facto Semantic Web language, to describe a map for a set of ontologies. The axioms will relate concepts from one ontology to the other. ... There is a well-established body of research in the area of automated ontology alignment. This is not our focus. Instead we investigate the application of these alignments to provide an integrated view of the Semantic Web data.

(emphasis mine). The rest of the paper justifies this approach, leading up to:

We first query the knowledge base from the perspective of each of the 10 ontologies that define the concept Person. We now ask for all the instances of the concept Person. The results vary from 4 to 1,163,628. We then map the Person concept from all the ontologies to the Person concept defined in the FOAF ontology. We now issue the same query from the perspective of this map and we get 1,213,246 results. The results now encompass all the data sources that commit to these 10 ontologies. Note: a pair wise mapping would have taken 45 mapping axioms to establish this alignment instead of the 9 mapping axioms that we used. More importantly due to this network effect of the maps, by contributing just a single map, one will automatically get the benefit of all the data that is available in the network.

That's fantastic stuff.

We now pause for a word from Steve Lawrence; NEC Research Institute, to lament the lack of free online proceedings for AAAI: Articles freely available online are more highly cited. For greater impact and faster scientific progress, authors and publishers should aim to make research easy to access. OK, now back to the great paper...

Along the way, they give a definition of a knowledge function, K, that is remarkably similar to log:semantics from N3. They also define a commitment function that is basically the ontological closure pattern.

The approach to querying all this data is something they call DLDB, which comes from a paper they submitted to the ISWC Practical and Scalable Semantic Systems workshop. Darn! no full text proceedings online again. Ah... Jeff's pubs include a tech report version. To paraphrase: there's a table for each class and a table for each property that relates rows from the class tables. They use a DL reasoner to find subclass relationships, and they make views out of them. I have never seen this approach to RdfAndSql before; it sure looks promising. I wonder if we can integrate it into our dbview work somehow and perhaps into our truth-maintenance system in the TAMI project.

This wasn't the only work at AAAI on scalable, practical knowledge representation. I caught just a glance at some other papers at the conference that exploit wikipedia as a dataset in various algorithms. I hope to study those more.

I also ran into Ben Kuipers, whose Algernon and Access-Limited Logic has long appealed to me as an approach to reasoning that might work well when scaled up to Semantic Web data sets. That work is mostly on hold; we started talking about getting it going again, but didn't get very far into the conversation. I hope to pick that up again soon.

I gather the 1.0 release of OpenCyc happened at the conference; there's a lot of great stuff in cyc, but only time will tell how well it will integrate with other Semantic Web stuff.

Meanwhile, a handy citation for Heflin's paper...

• Z. Pan and A. Qasem and J. Heflin An Investigation into the Feasibility of the Semantic Web. In Proc. of the Twenty First National Conference on Artificial Intelligence (AAAI 2006), Boston, USA, 2006 (abstract)

That's marked up using an XHTML/LaText/BibTex idiom that I'm working on so that we get BibTex for free:

@inproceedings{pan06a,
title = "{An Investigation into the Feasibility of the Semantic Web}",
    author = {Z. Pan and A. Qasem and J. Heflin},
    booktitle = {Proc. of the Twenty First  National Conference on Artificial Intelligence  (AAAI 2006)},
    year = {2006},
    address = {Boston, USA},
}

Wikimania 2006 was last week in Boston; I had it on my travel schedule, tenatively, months in advance, but I didn't really come up with a solid justification, and there were conflicts, so I ended up not going.

I was very interested to see the online participation options, but I didn't get my hopes up too high, because I know that ConnectingAudiences is challenging.

I tried to participate in the transcription stuff real-time; installation of the goby collaborative editor went smoothly enough (it looks like an interesting alternative to SubEthaEdit, though it's client/server, not peer-to-peer; they're talking about switching to the jabber protocol...) but I couldn't seem to connect to any sessions while people were active in them.

The real-time video feed of mako on a definition of Freedom was surprisingly good, though I couldn't give it my full attention during the work day. I didn't understand the problem he was speaking to (isn't GFDL good enough?) until I listened to Lessig on Free Culture and realized that CC share-alike and GFDL don't interoperate. (Yet another reason to keep the test of independent invention in mind at all times.)

Lessig read this quote, but only referred to the author using a photo that I couldn't see via the audio feed; when I looked it up, I realized there was a gap in this student's free culture education:

If we don't want to live in a jungle, we must change our attitudes. We must start sending the message that a good citizen is one who cooperates when appropriate, not one who is successful at taking from others.

RMS, 1992

These sessions on the wikipedia process look particularly interesting; I hope to find time to see or listen to a recording:

• The Process of Requests for Adminship on the English Wikipedia: The Role of Trust in an Open System
• A Question-and-answer session with the English Wikipedia Arbitration Committee

I bumped into TimBL online and remind him about the Wikipedia and the Semantic Web panel; he had turned it down because of other travel obligations, but he just managed to stop by after all. I hope it went allright; he was pretty jet-lagged.

I see WikiSym 2006 coming up August 21-23, 2006 in Odense, Denmark. I'm not sure I can find justification to make travel plans on just a few weeks of notice. But Denny's hottest conference ever item burns like salt in an open wound and motivates me to give it a try. It looks like the SweetWiki folks, who participate in the GRDDL WG, will be there; that's the start of a justification...

The developer track at WWW2006 last week in Edinburgh was really cool; you had to show up on time or you couldn't fit in the room! One of the coolest talks was D2R-Server - Publishing Relational Databases on the Web as SPARQL-Endpoints.. I see D2R Server is released now. Cool.

Yes, storing RDF in a SQL database using 3-column tables (or 4 or 5 or 6...) is cool as far as it goes, but I'm gland we're finally seeing more work on taking existing SQL databases (whose schemas are not designed with RDF in mind) and exporting them as RDF.

TimBL wrote a design note on Relational Databases on the Semantic Web in 1998. In 2002, I wrote dbview.py, a couple hundred lines of python that implements parts of it. Rob Crowell picked it up and the 2005/2006 version of dbview.py now does foreign keys and backlinks.

D2R gets points for using RDF for their configuration/mapping info. The slides showed turtle/n3. Why are the dbin brainlets in XML but not RDF? I wonder.

D2R Server has a mapping layer; dbview assumes that will be handled with rules. The choice of URIs for column names is interesting. D2R uses jdbc:mysql://127.0.0.1/wordpress#users1, but dbview is all about embedding a SQL database in HTTP space, so we use URIs like http://db.example/orders/customers/custno/1#item. In dbview, the decisions about when to use / and when to use # are made so that the result is browseable. In D2R, the default URIs don't matter as much because it's expected that they'll be mapped to a more well-known ontology/schema like foaf.

dbview is still just a few hundred lines of python; we haven't integrated the SPARQL parser that Yosi developed for cwm, nor integrated EricP's work on federated query.

Speaking of federated query... on Wednesday at the conference, I saw Tim Finin in the poster session. He showed me something the swoogle folks are cooking up: you give it a SPARQL query, and it looks at the terms used in your query and suggests documents you should put in your SPARQL dataset to run your query against. I hope to hear more about that.

Somewhere in EricP's work is one of the several SPARQL-to-SQL rewriters out there... oh... I thought the HP tech report, A relational algebra for SPARQL was another one, but it seems to be by Richard Cyganiak, one of the D2R guys.

Benjamin Nowack's Feb 2006 item announced a SPARQL-to-SQL rewriter for his ARC RDF store for PHP.

Hmm... maybe it's time for a ScheduledTopicChat on SPARQL, SQL, and RDF? If you're interested, suggest a couple times that would be good for you in a comment or in mail to me and a public archive.

I went to Edinburgh last week for WWW2006.

I spent Tuesday in the workshop on Identity, Reference, and the Web (IRW2006). I didn't really finish my presentation slides in time, but I think my paper, A Pragmatic Theory of Reference for the Web is mostly coherent. Each section of the workshop got an entry in a semantic wiki; mine is the one that started at 12:00.

The IRE formalism presented by Valentina and Aldo was though-provoking. I think their proxy-for is like foaf:topic (modulo the way they mix in time). And exact-proxy-for is like foaf:primaryTopic. Very handy. I wonder if foaf:primaryTopic should be promoted to its own thing, separate from all the social networking stuff in foaf.

Ginsberg's talk hit on one of the most important questions: "Do I commit to a document just because I use one of its terms?" His answer was basically to reify everything; I think we can do better than that. Peter Patel-Schneider's talk basically gave a 'no' answer to the question. I don't think we should go that far either, though from a standardization point of view, that's sorta where we're at.

Steve Pepper's talked about published subjects and public resource identifiers; I can sympathize with his point that we have too many URL/URI/URN/IRI/XRI/etc. terms, but when he suggests that the answer is to make a new one, I'm not sure I agree. He argues to deprecate all the others, but as URI Activity lead at W3C, I'm not in a position where I can overrule people and deprecate things that they say they want. I agree with him that the 303 redirection is too much trouble, but he doesn't seem to be willing to use the HashURI pattern either, and as I said in the advice section of my paper, that's asking for trouble.

On Thursday, I was on a panel about tagging versus the Semantic Web: Meaning on the Web: Evolution or Intelligent Design?. Frank started by debunking 4 myths about the Semantic Web. I gotta find Frank's slides. "I'll hold up one finger whever anybody says myth #1, and so on." As the the other Frank was talking about tagging, Frank held up 2 and 3 fingers, and the audience pointed out that he should have held up 1 finger.

I talked without slides. I think I got away with it. I said that I don't expect symbolic reasoning to beat statistical methods when it comes to the wisdom of crowds, but who wants to delegate their bank balance or the targets of their mail messages to the wisdom of crowds? Sometimes we mean exactly what we say, not just something close.

I suggested that GRDDL+microformats is a practical way to get lots of Semantic Web data. And I brought up the problem with iCalendar timezones and noted that while timezones data should be published by the government entities that govern them, Semantic Web data from wikipedia might be a more straightforward mechanism and might be just as democratic.

So much for philosophical discussions; stay tuned for another item about SPARQL and databases and running code.

Tue 15 Nov

• @09:00.Opening Keynote: From Atoms to OWLs the new ecology of the Semantic Web

  Jonathan Robie said Jim made 5 or 6 points in this talk that had been obscure, at best, in earlier talks on the Semantic Web in RDF. Kurt Cagle's notes don't seem to show them.

  Unfortunately, the DAWG teleconference started at 9:30 and missing it would have delayed the WG by several weeks, so I ducked out.

• @11:45.Handling Math in Real-World Workflows: Practical Lessons

  What jumped out at me was that primary school publishers are hot on MathML content markup for reasons of accessibility. It's good to hear that the theory that higher levels of semantic abstraction contribute to accessability plays out in practice.

  I asked if XSL-FO was on the map in this world of production math workflows, but he said no, not really.

  I wanted to ask what would you change about mathml if you had a time machine? but didn't find the right moment.

• @14:00.Modeling Methods and Artifacts for Crossing the Data/Document Divide
  
  KlaasBals XML 2005 - 9
  Originally uploaded by roland.

  He had this great slide (5 of 37) showing that business data goes from narrative documents (catalogs) to transaction data (orders) and back to narrative (support docs) and that data models need to cross them.

  An HL7 person asked how important is it to trace back from document to data model? which prompted me to add a slide to my talk to make the point about how GRDDL lets you get from narrative documents back to UML-like models.

• *14:00.Federated Identity Management: An Overview of Concepts and Standards

  I really wanted to get this overview from Eve, but I got caught in a hallway conversation or something and missed it. I picked up some of how Liberty works from the next talk. ID-FF looks an awful lot like OpenID. I wonder what's the difference.

• @14:45.Liberty Federation Deployment Case Study

  This showed real-world deployment of federated identity services remarkably like the ones we discuss in the PAW project. I asked a lot of questions about the details, and the answers were quite reasonable. Afterward, I said to Eve and Yvonne, Would you slow down? We're trying to pitch many of these ideas to research funders. If you deploy them all in commercial settings, where will we be? ;-)

  At some point, she mentioned government rules where authorization data was considered sensitive but authentication data was not. I hope to get more details about that.

• *14:45.Microsoft's Language Integrated Query and XML

  I heard this was a great talk; both the content and the presentation. I hope to get notes from Norm, Michael, and others who were there.

• @16:00.The Atom Publishing Protocol: Publishing Web Content with XML and HTTP

  Most of the stuff he presented looked familiar; looks like not much has changed since the last time I saw Joe give a talk on the Atom protocol. He mentioned great article by Udell on URIs... ah... Tangled in the threads seems to be a column that Udell used to write for Byte. Anybody got a pointer to the article Joe was talking about? I'd like to cite it among the TAG educational materials.

• @16:45.Remixing RSS - past, present and future

  An interesting perspective of the development of blogging. At their booth, I discovered bryght is a big drupal shop. I asked them about in-browser direct-manipulation editing; they're big on TinyMCE.

• *16:00 .Names, Namespaces, XML Languages and XML Definition Languages

  I think Henry alluded to this paper in a TAG discussion of XMLVersioning-41 in Edinburgh. I was looking forward to getting it presented conference-style, but I guess I can read the paper and discuss it in the TAG.

Wed 16 Nov

• * 11:45 . On Language Creation

  missed this in the panic of preparing for my talk. Darn.

• * 11:45 . Native XML Scripting with ECMAScript for XML (E4X)

  missed this in the panic of preparing for my talk. Darn.

• @ 14:00 . XSL Transform Self-Documentation
• * 16:00 . XML, REST, and SOAP at Yahoo

  Wanted to follow up on the conversation we started briefly in his blog but it was scheduled against a SPARQL session. Had to settle for a brief hand-shake and card exchange.

• @ 16:00 . SQL, XQuery, and SPARQL: What's wrong with this picture?

  What looked like a SPARQL-bashing session turned into a pretty good SPARQL tutorial. Jim Melton, who has been doing SQL standards work for over 20 years and sharing that experience in the XQuery WG for several years, took a close look at SPARQL, prompted by some nifty results by some folks using RDF/OWL in drug discovery. Even though he was "strongly encouraged" to conclude that SPARQL was obviated by SQL and XQuery, his conclusion was that it has a place.

Thu 17 Nov

• * 09:00 . Describing Web Applications

  missed the talk but spent some time noodling on WebDescriptionProposals in the ESW wiki. I hope to study WADL more closely.

• * 09:00 . Semantics and Security: Applying RDF and OWL to Defense and Security Challenges
• * 11:00 . The Impact of XML on Contract Law and the Volume of Contract Litigation

  I didn't get to see Jane Winn's talk (though it won an award and I look forward to reading the paper), but after Bob G. introduce me to her in the exhibit hall, we had a fascinating discussion of the social side of open source and open standards. There's some seminal paper on charismaric leadership that she's supposed to be sending me. I asked for a pointer, but it seems to come from the pre-Web world of paper and fax machines.

• @ 11:00 . Unit Testing in XSLT 2.0

  Some comments on XSLT 1.0 were pushed back a la but wer're just doing a transformation language for stylesheets, not a general-purpose programming language. Maybe so, but clearly XSLT 2.0 is sufficiently general purpose to build unit testing harnesses.

• @ 11:45 . Automated mass production of XSLT stylesheets

  Using a spreadsheet as a way to communicate design requirements and even decisions from users to developers and/or straight to the machine. Cute. Reminds me of my own work on using spreadsheets as an RDF authoring tool.

• * 14:00 . "Just" Use XML

  His slides have great stuff on pitfalls of XML, unicode, dateTime, etc.; filed it under quality

• @ 14:45 . A Generalized Grammar for Three-way XML Synchronization

  This was much less relevant to our work on RDF diff/sync than I thought it might be, but trick of using SVG animation to show differences between images was really cool.

• @ 16:00 . Using XSL, XForms and UBL together to create complex forms with visual fidelity

  I have missed so many chances to soak up XForms at a conference... I finally made the time for this one, but it turned out to me more about XSL-FO.

• @ 16:45 . Enterprise-level Web Form Applications with XForms and XFDL

  John Boyer is an XForms WG co-chair, as well as doing DSig stuff. He's now at IBM, since they acquired Pure Edge. Hendler got the two of us together in the hall to discuss some connection between XForms and SPARQL... something about using XForms in the role that XSLT plays in handling SPARQL results. Hmm.

  John asked that we "don't think of a Form as just the typical name/address or pizza order form"; likewise, he asked us to put aside our notion of table. But I didn't get a firm feel for what we're supposed to put in thier places, except that a form can be a blackjack game.

• @ 20:00 . Closing Keynote: Everyone's using XML, but does anyone care?

  Very entertaining. On Web services, he said we've got this little pea of xml under all these layers and only a true xml princess can feel it.