Categories
RDF Semantics

This Week’s Semantic Web, Burningbird style

Recovered from the Wayback Machine.

Last week, Danny Ayers made a request to the semantic web community at large: that we take turns publishing our own version of This Week’s Semantic Web. I volunteered to start, and hope that others follow, though in comments to Danny’s post, the suggestion about the Gem of the Week sounded better (and a lot less work).

However, I decided to add a slight twist to my own version of This Week’s Semantic Web, focusing not only on the stories, but how I found them. After all, the real purpose of the semantic web technologies is to make information easier to find. How are we, in the semantic web community, doing in this regard?

To start, I subscribe to various feeds including Planet RDF, as a way to keep up with most of the semantic web news. This week, the stories from Planet RDF that caught my eye were the following:

  • Tom Heath wrote How Will We Interact with the Web of Data for IEEE. In his article, Tom proposes that the web homepage, as we know it today, is dead. In its place we’ll have connected pieces of data, pulled together via RDF records (tuples), which are then used to generate the human readable content. So one could have weblog, browser, feeds, friend feeds, and other online “islands of data”, Flickr and other photos, videos on YouTube, etc.—all annotated with metadata and brought together, mechanistically, because of the metadata annotation. It’s interesting, and we already have some of this with various widget-enabled devices, but I’m not sure that most people are “geek enough” to make this a truly viable option. Not yet.
  • Bob DuCharme wrote a follow-up piece to his Leaning more about SPARQL, related to forming SPARQL queries against DBPedia, the site dedicated to making Wikipedia information queriable. No, that’s not a word…but it should be. Bob’s example is important for two reasons. The first, and the most obvious, reason is that it, of course, demonstrates SPARQL against a published source—hopefully spurring on other efforts. More importantly, though, in my opinion, is that Bob is publishing his explorations, his learning experiences, not necessarily a finished, “Ta da!” work. We need more journals of discovery in the semantic web world.

I don’t only get my semantic web information from the Planet RDF feed. I find other entries on this topic, now and again, in other feeds. For instance, I wrote about two other items this week and I’ll repeat links to both because I feel they represent the semantic world “in the wild”.

  • A List Apart featured an article titled Understanding Progressive Enhancement, which discussed the concept of building one’s website from the inside out—focusing on the properly semantically annotated content, first, before tossing in the pretties. I think this article complements some of the discussion about minimal design that was such a popular topic a few months back. The article not only focuses our attention back on the content, and hence the real purpose for the web site, it also drives home that we need to start doing a better job, semantically speaking, with our use of page markup. Speaking of markup…
  • Tina Holmboe’s XHTML—myths and realities is both an important, and timely, look at XHTML, the importance of XHTML for the semantic world (RDFa), and the future of XHTML. It’s timely because it serves to remind us that we now have two divergent markup paths under the W3C leadership—paths that do not share a common model or focus, which seems to me to act counter to the ultimate goal of a truly semantic web.

In my quest for this week’s semantic web goodies, I also searched in Google on “Semantic Web” and then focused on News, not Web, in order to filter items down to recent events. With this approach, I found the following items to pass along:

    • Paul Miller at ZDNet writes Does the Semantic web matter? He believes it does, a view offered up simply and elegantly. What the semantic web isn’t, though, according to Paul, is a goose to be punched and pummeled by the elitist and the avaricious until forced to deliver up the golden egg. To wit:

Continuing landgrabs by startups that seek to attract, trap and exploit eyeballs stand unashamedly on the shoulders of Semantic Web promise whilst running counter to its basic tenets of linking and openness. On the other hand, companies ‘just’ doing perfectly reasonable – and valuable – things with the meanings of words, phrases and documents latch on to the Semantic Web’s buzz, whilst being all about Semantics and not at all about the Web.

New entrants, hopefully building viable and useful businesses upon the Semantic Web’s ideas, are pilloried by stalwarts of the ‘community,’ because the reality of their business model does not permit a whole-hearted embracing of the entire Semantic Web stack from Day One. Intellectual purity clashes with pragmatism and reality on a daily basis. Well-meaning guidelines and best practices morph in the minds of too many to become laws, ‘truths’, and rods with which to beat outsiders. Visions of Orwellian pigs fill my brain, and I don’t like what I see as they rise up onto two feet and gaze disdainfully around.

  • Speaking of punching geese, oh look, Ask.com is back. It’s got mad semantic skillz. So I put Ask.com to the test, and asked it “How can I learn more about SPARQL”, and it responded with, “Did you mean, ‘How can I learn more about sparkle’?”. I paused a moment, and said sure, show me that one. Ummm, Swarovski crystal jewelry. Pretty sparkles. To be fair, before following this sparkly tangent, Ask.com did return the first of Bob Ducharme’s post, mentioned above. In fact, it returned exactly the same result list as Google and Yahoo, when I asked them the same question.
  • Though not exactly “this week”, ReadWriteWeb writes a mean semantic web post, now and again, and had one last week subtitled, “Show me the Money!”—and wasn’t that a great movie moment? I digress, though. The RWW post focuses on a new report by a Semantic web entrepreneur on semantic web companies making money, but just at the moment when I clicked through to read the report, I got distracted by the flock of migrating geese overhead. I must pursue the report at a later time. What I found interesting, though, was the ReadWriteWeb Semantic Web Log search and…ah geez, there goes another flock, circling overhead.

There were other sources I searched for information about the semantic web for this week, but the results were less than optimum. For instance, I searched on “semanticweb” in delicious, but the results show the items that were posted to delicious this week, not necessarily published this week. The problem is that while many services such as delicious have a way to tag items with terms like “semanticweb” the metadata annotation is limited, and doesn’t include information such as when was the posted item first published, nor allow you to search on the same. Most of the “semantics” are flat, simple, and two-dimensional, IE keyword-value pairs.

I next went in the opposite direction, looking for just published items, and then sought to filter on the semantic web. For instance, no other source is better for up-to-date discovery of minutiae than Twitter. However, as far as I can see, there is no way to search on specific topic in Twitter. You can look for people, but other subject material search is extremely limited. If you don’t know that Twitter user Kingsley Idehen exists, and posts frequently on semantic web related items, you may not discover a graph of linked data sources or an animation related to RDF as middleware.

I then turned to the Big Cheese, the Head Semantic Web honcho, Twine, and the twine related to the Semantic Web. Eureka! I finded the Semantic Web! Of course, on closer look, most of the items also could be found on Planet RDF. Still, meat that is both fresh, and relevant. I’ll just pick out a few for my version of This Week’s Semantic Web.

  • Seven OWL 2 Drafts Published at the W3C. OWL 2 is an extension of the OWL, which is the Web Ontology Language. No, don’t try to fit the acronym. OWL is not necessarily directly important to thee and me. OWL is important, though, for designing systems that would understand exactly what I mean when I ask, “How can I learn more about SPARQL”, and that will return the definitive sources meeting my question, without being dependent on either language processing or obscure page ranking algorithms.
  • Speaking of SPARQL, another item in the twine was SPARQL Update a submission to the W3C describing a way to use SPARQL to update graphs (semantically linked data stores). Interesting, considering that SPARQL means Simple Protocol and RDF Query Language. What works in one direction must work all directions, eh? Reminds me a little of HTML5 and JSON—the Swiss Army knives of technology.

And so ends my tenure for This Week’s Semantic Web, Burningbird style. What I discovered in the process of building my list was that we’re not close to the semantic web we seek. Without knowing about the people, such as Bob, Kingsley, or Danny, or the topic-focused resources such as Planet RDF or Twine, I would have had a much more difficult time finding out what is happening, this week, in the semantic web. However, among the results I did find are new technologies, new specifications, new efforts that assure us that though the semantic web doesn’t exist today, it surely will someday.

Surely. Someday.

flack of snow geese


update

Brian Manley also accepted Danny’s challenge with This Week in Linked Data.