The Tetherless World Weblog

I wish I had more time to write a real blog on this, and I know most of these have already been announced in various longer blogs by others - but in case you’ve missed it, the past week or so has seen some real indicators of Semantic Web growth. Obviously it is good that Microsoft is hosting a Palo Alto Semantic Web meetup (http://www.powerset.com/blog/articles/2009/03/19/call-for-presenters-epic-paws-meetup) and the Semantic Search announcement by Google can’t hurt (cf. http://www.fool.com/investing/high-growth/2009/03/25/its-all-semantics-says-google.aspx), but here’s some other “we’re getting real” announcements that have come out recently:

1 - announcement of Obama use of Sem Web: http://www.sitepoint.com/blogs/2009/03/19/obama-groundbreaking-use-semantic-web/
2 - announcement of “Semantic Web for Dummies” - http://www.amazon.com/Semantic-Web-Dummies-Computer-Tech/dp/0470396792
3 - announcement of Oreilly “programming the semantic web” - http://oreilly.com/catalog/9780596802066/
I’ve been told that when you’re in a community, you cannot really see the exponential growth thing until “n” gets large enough — I don’t know for sure, but I’m feeling like we may be getting to that “n.” So for folks like me who’ve been doing so much evangelizing for a while now, it’s almost scary — when I was at DARPA, back in those days when the Semantic Web was a twinkle in Tim BL’s eye, I used to say that someday there would be weeks like this - but I don’t know if I ever really believed it. Now, I think I do…

Way to go Web 3.0,

Jim H.

p.s. and of course, this also means my research group better start figuring out what Web 4.0 is going to be, because this 3.0 space is starting to get crowded…

Stefan Decker raised a case “missing children” indicating that Description Logic, or more precisely open world assumption, maybe overkill because it is not needed in most real world applications even though AI researchers like it.

Michel Dumontier, as a researcher trying to adopt OWL, reviews the well-known semantic web benefits, again, from the SW developer community who hopes these nice features will be helpful to real world Web developers and users.

Tim Finin, “maybe we’re a victim of our own success”. Moving towards KR monoculture could be quite dangerous. He raises some examples that OWL does not fit, e.g. when encoding knowledge extracted from knowledge, a lot of information lots such as time, uncertainty, and provenance.

Ian Horrocks claims that OWL has good connection with other KR approaches, OWL is not going to solve all problems, but it is useful in general.

One biggest argument raised by many is that “is OWL useful?” not even “is OWL2 useful”. Of course there are both supportive and negative evidences, and neither side can convince the other side. Someone also argue that the learning curve of OWL will just stop potential user. (Industrial adoption is a better benchmark because researchers are more flexible).

Another issue is “scalability”. Jim Hendler tried to be even worse than Stefan, Twine is claimed to Semantic Web applications, it use a few pieces of OWL to scale up. In general, scalability is the non-negotiable requirement of Web data computation. That is database community avoids, for instance, recursion in relation algebra.

A third question, raised by David Karger, “what are we doing with OWL? Which pieces of OWL are actually being used, and Why?” (This is actually a motivation for OWL2, and why three OWL2 fragments are proposed. We are looking forward to see if OWL working group can give industrial users a good answer.)

Well, a fourth question is “OWL2 is KR, i.e. a family of Description Logics profiles that link to other KR languages?” and/or “OWL2 is trying to promote better web or semantic web applications?”.

Closing remarks (I did my best to keep it original)
* Ian, “choose hope, not fear”
* Tim, “I can see Russia from my house”
* Michel, “OWL is pretty good language”
* Stefan, “if you did not fix the little thing, you may miss the boat”

By Li Ding
Greetings from ISWC 2008

I just noticed the “OWL 2 Web Ontology Language: Requirements” document from the OWL Working Group. Interestingly, while the “W” in OWL stands for “Web”, I didn’t see any use case from web applications in the usual sense. As the leading requirements are from the need for domain knowledge bases, I would suggest the name of the new language, instead of OWL 2, to be Ontology Language of Domains (OLD) — Just kidding. OWL claims to be needed by common web users, but such users are surprisingly under-represented in the specification process. We have already seen many specially designed, highly expressive, but, narrowly applied languages in the old KR schools. Do we need to invent yet another one here, again?

Jie

The Tetherless World invites users of Apple’s iPhone and iPod Touch to try out our new ontology browser, OWL Mobile.

OWL Mobile is powered by Jena and Pellet, operating remotely, to provide speed and battery performance mobile devices users expect from their applications. Load one or more ontologies through the Load Ontologies tab. Supply a URL to a custom ontology or use the list of past ontologies. Once you’ve loaded an ontology, use the “Classes”, “Properties”, and “Individuals” tabs to browse through the ontology. Clicking on an item will expand it and give additional information about that particular object. Links which point to other members of the ontology will switch to the appropriate URI when clicked. External links such as web pages, email address, and phone numbers will open the appropriate application on iPhone (phone numbers won’t work on the iTouch) when activated.

Point Safari to http://onto.rpi.edu/demo/owlmobile2/ to try the application. Feel free to bookmark it or add it to the home screen for easy access.

Evan Patton

Summary — there is a new “profile” of OWL Full that might be of great interest to the RDF/Data Web community — read on:

To those who follow W3C happenings, you know that I’ve had some problems with, and resigned from, the new OWL Working Group. The problems have mainly been related to the philosophy of what this is all about, more than the details of specific language features, and maybe I’ll blog about that some other time. However, in this entry I want to say something positive about one small piece of what the working group has done, and direct the RDF community to take a look at it– I believe it may be close to something we’ve needed for a long time.

In the “OWL 2 Web Ontology Language: Profiles” document (http://www.w3.org/TR/2008/WD-owl2-profiles-20080411/) the group has created a new set of OWL profiles (formerly called fragments) so instead of OWL Lite, DL, and Full, we now have (probably to be renamed at a later date) OWL 2 Full and a number of profiles OWL 2 DL, OWL 2 EL++, OWL 2 DL-Lite, OWL 2 R DL, and OWL 2 R Full (there are also be the unnamed RDF equivalents of the EL++ and OWL DL-Lite, but the group refuses to acknolwedge that, a primary reason for my leaving — but that’s another story again).

Anyway, it is to the last of these “OWL 2 R Full” that I would like to direct the attention of the RDF community — it is a bit hard to tell from the relatively cryptic document, but this fragment is an extension to RDFS that adds a small amount of useful OWL vocabulary, without requiring commitment to some of the strong restrictions needed for the various DL dialects. The specification includes an axiomatic specification of the language (i.e. rules) and starting to circulate, but not in the OWL group’s document, is an N3 version of the language making it very easy to see the relation to RDF. A couple of the larger members of the Working Group have stated that they will support this language (I’m not sure whether in public or not, so I’ll let them speak for themselves) which bodes well.

For those people looking at the “Data Web” or at “Web 3.0″ applications, I think this profile of OWL may be worth looking at — it would definitely be improved by some comments from serious Web 3.0 application developers - as it may well be a good target of opportunity for further RDF development. In the famous Semantic Web layercake, this profile (which I would like to see renamed RDFS 3.0) would be able to sit under the Rules and Ontology fragments, where RDFS is now, without derailing RDF(S) into the peculiarities of description logics, yet allowing some useful constructs to be added. For example, FOAF, DOAP and other of the most used RDF-based ontologies would be within (or close to) this new profile

So if you’re not interested, or are studiously ignoring, the OWL drafts, let me suggest you take a look at Table 2 of section 4 of the Profiles document (and section 4.2.3 if you want to see the rules). I also suggest that one does not have to understand anything else in that section (much of which seems to me to be written for those with PhDs in AI or similar background) to be able to see there’s something useful in here.

So take a look at OWL 2 R Full - the name is awful, but the language might be a really powerful new tool on the RDF Web.

-Jim Hendler

p.s. Let me also suggest taking a look at the public email by Michael Schneider at http://lists.w3.org/Archives/Public/public-owl-wg/2008Apr/0171.html – one of the few RDF proponents in the working group, he gives a great example of using OWL R Full in an RDFS context…