Slashdot Mirror

← Back to Stories (view on slashdot.org)

Semantic Web Gathers Substance

Posted by timothy on from the an-as-yet-unidentified-substance dept.

David Hersey writes "ADTMag reports that the semantic web technologies are taking real form in the wake of recent W3C approvals and early pioneering work by vendors such as IBM, Boeing, Adobe and others. These technologies have been developing for several years. When and If the finally take form, they hold the potential of raising the capabilties of internet users and internet technology applications to levels that are today impractical due to the web's document-centric architecture."

26 comments

  1. Godel's Theorem? by ClosedSource · · Score: 1

    I heard that the Semantic Web violates Godel's Theorem but I didn't get any details. Any comments?

    1. Re:Godel's Theorem? by TRACK-YOUR-POSITION · · Score: 2, Insightful
      I never heard of that, but I suppose that saying a Semantic Web is in principle capable of more than the "Regular Web" probably violates something Turing said. It seems to me that XML is poorly designed to serve a marginally useful purpose. Why the hell does it have to look like HTML? Instead of all those starting and closing tags tripling the size of my documents, why not just use white space and parentheses or brackets so that it could have maintained some semblance of still being human readable?

      I'm not buying into the magical benefits of XML at all. Just because a file is written in XML doesn't mean your computer magically knows what it means (if that's what the semantic web is, then yeah, that violates both Godel's Theorem and common sense.) Your computer still has to know what XML tag to expect where, and what the data actually means in every tag (they can throw vague, poorly understood words like "ontology" around all they want, at the end of the day a human programmer still has to look up the documentation to see what a given tag stands for.)

      Perhaps using an XML parsing library is slightly easier than designing your own file format and documenting it properly (you have to document it either way...but XML makes the process more formal, in theory). Perhaps using an XML editor or viewer to edit XML files is slightly easier than using emacs to edit text files. I suspect the minimal efficiency gains from either of those have already been burned away by millions of people trying to read and sort through meaningless noise-philosophies by MBAs preaching the benefits of this New World Order of Ontological Semantic Standardization.

    2. Re:Godel's Theorem? by Uzziel · · Score: 3, Interesting
      You're not giving us a lot to go on here. How does it violate the Incompleteness Theorem?

      The semantic web is not a procedural system; it's a method of encoding information. Goedel's Incompleteness Theorem is about mathematical systems and their ability to describe certain truth values.

      I don't see any immediate connection between the one and the other.

    3. Re:Godel's Theorem? by Asmodeus · · Score: 1

      "Why the hell does it have to look like HTML? Instead of all those starting and closing tags tripling the size of my documents, why not just use white space and parentheses or brackets so that it could have maintained some semblance of still being human readable? "

      Like lisp S-Expressions?

      (tag 'data "other data" (sub-tag (list a b c)))

      Asmo

    4. Re:Godel's Theorem? by ClosedSource · · Score: 1

      Well, I said I didn't have the details, which is why I asked the question in the first place.

      I seems to me that a system that encodes information and then performs operations on it can be considered a mathematical system, so I don't think the semantic web is necessarily free from the implications of Godel's theorm.

    5. Re:Godel's Theorem? by thoth39 · · Score: 1

      I don't understand the problem with XML.
      I mean, if this format of data-storing was that bad, why the hell is SGML being used to this day?

      Anyway, the point of XML is not to automagically be meaningful for every computer software, it is to be parseable by every computer software. At least, for every XML parser computer software.

      It means you will be able to feed it into a DOM tree and analyze it no matter what schema it conforms to -- no matter how many schemas it conforms to either -- hell, if you're nice to it, you may even find *parts* of it you (the software) can understand. This is achieved by the single fact that the document is XML. Nevermind, at first, if it is valid.

      Based on this fact, you may have documents on the Web that carry semantically meaningful information for a variety of "clients" that are still readable by other clients without needing a parser full of special-exceptions-found-in-practice quirks mode.

      With that in mind, people are encouraged to add semantic information to their (XHTML or whatever) pages, knowing that at the very least, it will be ignored. Slowly, this will increase, and "semantic web enable" browsers will offer some neat stuff based on it. Like that Mozilla extension that verifies it a page is licensed under a Creative Commons license.

      --
      -- Pedro
    6. Re:Godel's Theorem? by Anonymous Coward · · Score: 1, Insightful
      I'm not buying into the magical benefits of XML at all. Just because a file is written in XML doesn't mean your computer magically knows what it means (if that's what the semantic web is, then yeah, that violates both Godel's Theorem and common sense.)
      Well obviously, but if you've got a ASCII file format vs a binary one - which do you think a developer will find easier to understand?

      XML is just the next step above ASCII. It's some simple grammar. It's hyped -- but it's still a good step.

      Perhaps using an XML parsing library is slightly easier than designing your own file format and documenting it properly (you have to document it either way...but XML makes the process more formal, in theory)
      There was this quote on Slashdot a few years ago saying that when a decision doesn't really matter -- when you could do it your own way or via standards, you should choose standards.

      Although designing your own format is quite possible, making your own parser isn't, and enforcing syntax isn't (schema/rest), and editors, and I very much doubt if you're going to have a stream-based parser that can process different parts of your syntax simultaneously ;)

  2. For some examples... by tcopeland · · Score: 4, Informative

    ...check out various tools on SemWebCentral.

    For example, you can browse the GForge project listing using OWL - more precisely, using an HTMLized version of the ObjectViewer.

    --
    The Army reading list
  3. Is this like the semantic antivirus? by Anonymous Coward · · Score: 0

    I keep getting semantic antivirus offers via email. Will I get similar offers for their web product next?

  4. Information architechure by heldlikesound · · Score: 2, Interesting

    IA is a new interest of mine, specifically on the web...

    My favorite sites are:

    boxesandarrows.com
    zeldman.com
    alistapart.com
    http://argus-acia.com/white_papers/iaglossary.htm l
    http://iaslash.org/

    --


    Cloud City Digital: DVD Production at its cheapest/finest
  5. Viability by !3ren · · Score: 5, Interesting

    Ultimately though, the question is whether people actually _want_ to enable the ability to reuse and recombine their data.
    Within your own data, tools to enable you to analyse and reuse your data are highly prized, but enabling anyone else to use your valuable data to their own ends benefits (almost) noone. (IMHO)

  6. Norton by Space · · Score: 0, Flamebait

    I hope the people that own Norton dont try to pursue any legal action about this due to the name similarities ;)

    --
    I Don't Work Here
  7. I thought it said Semitic Web by Anonymous Coward · · Score: 0, Funny

    People keep telling me we Jews run the World,(which is a silly notion) but I thought maybe we had taken a step forward.

  8. hmmm, seems fragile by jkorty · · Score: 3, Insightful

    The Semantic Web seems to be a generalized version of metatagging combined with a search engine tailored to the format. To work well, this requires 1) everyone to think hard and attach to their pages the appropriate semantics, and 2) that there are few people in the world that deliberately associate all sorts of junk with their web pages in order to get the page to appear in everyone's semantic search.

    1. Re:hmmm, seems fragile by BorgCopyeditor · · Score: 3, Insightful
      To work well, this requires 1) everyone to think hard and attach to their pages the appropriate semantics,

      Exactly. Given that people seem largely unable even to mark up documents with HTML properly, my hopes for the proper implementation of a significantly more "robust" ontology are somewhat dim.

      --
      Shop as usual. And avoid panic buying.
    2. Re:hmmm, seems fragile by Anonymous Coward · · Score: 0
      try explaining that to the guys leading semantic web and the response you will get is "do it our way or not at all."

      there have been other articles in the past pointing out these flaws. One of the important concepts of the current semantic web is an URI should and must be authorative. Obviously that doesn't happen in the real world. the other major flaw is the current approach uses model theory + monotonic logic. Although plenty of people have pointed out that approach won't scale, the semantic web lead refuses to listen. Oh well.

    3. Re:hmmm, seems fragile by anomalous+cohort · · Score: 1

      The ubiquitousness of OWL/RDF on the web would, IMHO, be tied to two things; adoption by a large search engine such as google and the availability of a cool tool that would make the authoring process easy.

      Imagine a mind mapping tool like freemind that could export to OWL. That would work for static content. Content publishers would do it for the same reason that they include keyword meta tags in their pages now.

    4. Re:hmmm, seems fragile by K-Man · · Score: 1
      The GIS community went through a similar thing with various metadata initiatives. People can put anything on a map, and the government tried to standardize its map layers into a centralized ontology, i.e. one type for streams, one for roads, one for pipes, etc. Unfortunately it became incredibly difficult to manage, and there were full-time data managers working to fit things together.

      To quote Ted Nelson:

      Intertwingularity is not generally acknowledged -- people keep pretending they can make things deeply hierarchical, categorizable and sequential when they can't. Everything is deeply intertwingled.
      I agree with Ted. I think the future is in better search tools, to handle ad-hoc queries instead of having people pre-designate their own work.

      --
      ---- "If we have to go on with these damned quantum jumps, then I'm sorry that I ever got involved" - Erwin Schrodinger
  9. Well by Anonymous Coward · · Score: 0

    It really depends on what you mean by 'semantic'.

  10. duh, INTRANETS by Ars-Fartsica · · Score: 1

    For the internet this will never take off, the disparate sites on the web will never grok it in a unified fashion. In Intranets though there can be goals, policies, centralized design and predetermined semantics. In these environments this type of stuff makes sense - there is a need and a return on investment.

  11. It seems like a meaningless concept by K-Man · · Score: 2, Insightful

    So far the biggest semantic web project I've seen is IBM's WebFountain, which is basically a big sed script that goes through the web and wraps each stock phrase it finds with meta tags, and enters them in a big database. It seems like a reasonable phrase search would accomplish the same thing.

    --
    ---- "If we have to go on with these damned quantum jumps, then I'm sorry that I ever got involved" - Erwin Schrodinger
  12. Not very realistic by Frans+Faase · · Score: 2, Insightful

    In a sense I am a great fan of the idea of a semantic web. But I am affraid that it does not work in reality. As soons as it gains some momentum it will be hyjacked by large companies (MS, Yahoo) trying to commercialize it and by small companies trying to misuse it (spam).

  13. Most people don't get it yet. by metoc · · Score: 1

    The concept of the Semantic Web is still in its infancy. Like many technologies before (XML, HTTP, Open Source, the "Internet", etc.) it will be a few more years before it becomes obvious.

    The main difference between the semantic web and web pages with meta-tags is that the current approach gets you hundreds of useless hits on google. You should by able to say "I want to fly to Whistler" and not get references to zippers, paintings and insects. The semantic web will allow you to do that.

    1. Re:Most people don't get it yet. by shivanan · · Score: 1

      You should by able to say "I want to fly to Whistler" and not get references to zippers, paintings and insects. The semantic web will allow you to do that.
      As someone previously posted, this mandates that people actually put meaningfull (read: not misleading) information in their meta data tags.
      I always figured the main advantage of the semantic web was that by providing data in machine-parseable format, you can have the computer automate a lot of stuff for you.
      I.e, "book tickets for me on the earliest flight that goes to New York." should make a program check flight listings, fetch the relevant result and send a request to the ticket server.
      This could theoretically be done by grokking through HTML pages but since since HTML design is non-standardized, you'd have to roll out a screen scraper for every page. The Semantic Web should standardize all of this.

  14. Impractical? by Nippoo · · Score: 1

    Why do they say "Impractical"? Why would id be impractical to have a semantic web?