Linked Data? Web of Data? Semantic Web? WTF?

This post was prompted by this tweet from Tim O’Reilly

People learning about Linked Data frequently ask “what’s the relationship between Linked Data and the Semantic Web?”, which is a fair and good question. One of the responses that crops up relatively frequently is that Linked Data is just an attempt to rebrand the Semantic Web. In my experience these kind of rebranding comments come mostly from people who have a certain impression of the Semantic Web vision (which may or may not be accurate), don’t like this vision, and therefore dismiss Linked Data on this basis without actually considering what it means (i.e. a means to dismantle data silos), and without necessarily rethinking their original view of the Semantic Web concept. I prefer to see it this way…

Think about HTML documents; when people started weaving these together with hyperlinks we got a Web of documents. Now think about data. When people started weaving individual bits of data together with RDF triples (that expressed the relationship between these bits of data) we saw the emergence of a Web of data. Linked Data is no more complex than this – connecting related data across the Web using URIs, HTTP and RDF. Of course there are many ways to have linked data, but in common usage Linked Data refers to the principles set out by Tim Berners-Lee in 2006.

So if we link data together using Web technologies, and according to these principles, the result is a Web of data. Personally I use the term Web of data largely interchangeably with the term Semantic Web, although not everyone in the Semantic Web world would agree with this. The precise term I use depends on the audience. With Semantic Web geeks I say Semantic Web, with others I tend to say Web of data – it’s not about rebranding, it’s about using terms that make sense to your audience, and Web of data speaks to people much more clearly than Semantic Web. Similarly, Linked Data isn’t about rebranding the Semantic Web, it’s about clarifying its fundamentals.

Tim Berners-Lee said several times last year, in public, that “Linked Data is the Semantic Web done right” (e.g. see these slides from Linked Data Planet in New York), and who am I to argue, it’s his vision. But to see this as a recent trend or a u-turn ignores the historical context. On page 191 of my copy of Weaving the Web (dated 2000, ISBN-13: 9781587990182) it says:

The first step is putting data on the Web in a form that machines can naturally understand, or converting it to that form. This creates what I call a Semantic Web – a web of data that can be processed directly or indirectly by machines.

I’m not sure this quote adequately captures the importance of links in the whole picture, but no one can claim that the Web of data label is recent marketing spin invented to make the Semantic Web palatable. This was always the deal. It’s certainly how I understood the concept (and what inspired me to do a PhD in the area).

If others  have somehow diverted the Semantic Web vision down some side road since Weaving the Web was written, then that’s unfortunate. (In my experience the Linking Open Data project was an attempt to reconnect the Semantic Web community with the some of the key aspects of the original vision that were being overlooked, like having a real Web of data as the basis for research). I certainly notice plenty of unjustified attempts at present to co-opt the term Semantic Web, now that it’s no longer a dirty word, and drive it off down some dodgy alleyway. Some of these products, services or companies may be applications or services that use some semantic technology and are delivered over the Web, but that doesn’t make them Semantic Web applications, services or companies. Anything claiming the Semantic Web label needs to get its hands dirty with Linked Data somewhere along the way. That’s just how it is.

So to return to Tim O’Reilly’s tweet, he’s not far wrong about the lack of difference between Linked Data, Semantic Web and RDF (we’ll ignore the means vs end vs technology distinction), but I’d love to know who he’s quoting about the explicit rebranding.

29 Responses to “Linked Data? Web of Data? Semantic Web? WTF?”


  1. Greg Boutin

    Nice illustration of the “TBL follower camp” I described in this post, based on the discussion here

    “Anything claiming the Semantic Web label needs to get its hands dirty with Linked Data somewhere along the way. That’s just how it is.”

    Tom, TBL’s vision may be attractive, but it’s not the word of god, and there are other ways to get machine to leverage the meaning in the data than to build that meaning directly in the data. Some of them algorithmic, others as alternative attempts at standardizing metadata and “smarter linking”. Even if it is making nice progress, Linked Data is not a forced passage in this massive undertaking, and assuming all apps that don’t do linked data are misusing the SW appellation appears to me as very top-down, quasi-fundamentalist approach.

    “Some of these products, services or companies may be applications or services that use some semantic technology and are delivered over the Web, but that doesn’t make them Semantic Web applications, services or companies.” I’d say it does, actually. It just doesn’t make them Linked Data.

  2. Tom Heath

    Hi Greg.

    You say this:

    there are other ways to get machine to leverage the meaning in the data than to build that meaning directly in the data.

    None of what I’ve said contradicts this. Of course it’s true! The key factor here is what happens when the owner of that machine wants to expose the results of that process for others to consume, possibly in ways that the machine owner cannot anticipate. It’s here that the Semantic Web technology stack is particularly important, as it provides standardised ways of doing this.

    Of course special arrangements can be made for each third party consumer, but that would be a bit like mailing a catalogue out to everyone who may want to buy your products, rather than just letting them look at your Web site.

    There is huge scope for backend machine intelligence to improve applications and services, of course, and not all the results of these processes will need to be exposed to other applications, but where they do, then the Semantic Web technology stack in my opinion represents the best technical approach to avoiding these being silo’d.

    There is a potentially infinite lifecycle here of publication, consumption, processing, publication, consumption, etc. I get the impression you’re most interested in the processing side of things, which is fine, it’s where the real intelligence will happen. I happen to have been focused on the publication and consumption bits recently, but none of these are at odds with each other. In fact the more of each we have the better.

    It seems we disagree about what constitutes a Semantic Web application. I’m not going to spend time here trying to persuade you to change your opinion. However, given that the term Semantic Web (deliberate uppercase) is widely associated with a particular technology stack, would it not make sense for the kinds of applications and services you refer to that don’t do Linked Data (publication or consumption) to go by a different label, say Smart Web or Intelligent Web? That would probably help clarify things for many people.

  3. Lee Feigenbaum

    Tom, I must say that I find the attempt to conflate the Semantic Web stack of technologies with Linked Data a bit alarming.

    I attended the C-SHALS conference in Cambridge, MA the past few days, and one of the observations that I made during a discussion session was that one of the least productive–yet very regular–points of discourse in that community is whether one application or another is the “right” way to use ontologies. My thoughts was that that’s a pretty destructive way to view things, and instead we should acknowledge that we’re drawn together by a common core set of technologies and that the “right” way to use those technologies should very much be driven by the business case / use case in question. Sometimes it’s “right” to use ontologies for modeling vocabularies or taxnomies; sometimes it’s right to use them as data schemas; sometimes it’s right to use them for classification & consistency checking, or for query answering. None is more right than the other.

    I feel similarly about what you’re doing here. You’ve (correctly, IMO) observed that a key fact about Semantic Web applications is that they’re built on the W3C stack of technologies. But then you’ve made an (unjustified, IMO) leap to say that therefore those applications must adhere to Linked Data principles, which are not part of that core W3C technology stack. (As one example, there’s no member-driven and community-reviewed W3C Recommendation for Linked Data.) And even if Linked Data were part of this stack, asserting that a Semantic Web application _must_ use one piece of the stack is somewhat anathema to the right-tool-for-the-right-job attitude of the rest of the stack that I and others find so appealing.

    There are great Semantic Web applications that use only OWL. There are others that use only RDF, but without HTTP URIs. There are some that use RDF and OWL, but don’t make resource identifiers dereferenceable. To write all those off “non-Semantic Web” really bothers me.

  4. Greg Boutin

    Hi Tom,

    You’re right, whether one uses a built-in-metadata vs. algorithmic approach, that doesn’t matter if it both gets machines to better understand the meaning of that data. My impression was that Linked Data advocates a more static approach, but I see how you could disagree and won’t try to argue that point.

    However, you miss the core argument I make, which is that linked data is not a requirement for the semantic web. It is one of the most successful elements so far, but it is not a forced passage. I think Lee, on your blog’s comments, explained it nicely.

    You propose using a “different label than semantic web” to describe semantic apps that don’t use linked data. On the other discussion I am a part of, they proposed “semantic technologies”. My response is that, beside being unrealistic b/c top-down definitional attempts like this rarely work, it also would be counterproductive b/c (1) the semantic web is a term that implies broader benefits than those linked data is capable of offering (2) it would be seen as an attempt by the linked data community to force-feed their technologies. Why not simply call apps that use linked data, “linked data apps”?

    Linked data is neither the semantic web nor should be seen as a forced passage. But ultimately, our differences hinges on whether you adhere to a strict definition of the SW, TBL-inspired, or a looser one, mass-market-inspired. I may be wrong, but I’d argue that the latter will prevail.

  5. Greg Boutin

    Mmm, rereading this, sorry for saying “you missed the core argument”, I see you didn’t, but we differ on it :)

  6. Tom Heath

    Hi Lee,

    …we should acknowledge that we’re drawn together by a common core set of technologies and that the “right” way to use those technologies should very much be driven by the business case / use case in question. Sometimes it’s “right” to use ontologies for modeling vocabularies or taxnomies; sometimes it’s right to use them as data schemas; sometimes it’s right to use them for classification & consistency checking, or for query answering. None is more right than the other.

    I agree, absolutely. (Although to slightly change the emphasis of your statement, I feel more drawn to making better use of information – through sharing, linking, reuse, etc – and happen to see the Semantic Web and its associated technology stack as a great facilitator of that).

    I don’t think I’ve said that there’s a right or wrong way to use these technologies, which (if I’ve interpreted your comments correctly) you seem to be suggesting, only that in my view there are certain prerequisites to something being classed as a Semantic Web application/service, and that right now many things that claim to be Semantic Web applications are not. That doesn’t make them wrong, it just doesn’t make them Semantic Web applications. They may make brilliant use of parts of the Semantic Web technology stack, and still not be Semantic Web applications. By way of analogy, I could have a purely desktop application that uses HTML for document markup and CSS for presentation, but this wouldn’t make it a Web application.

    There are great Semantic Web applications that use only OWL. There are others that use only RDF, but without HTTP URIs. There are some that use RDF and OWL, but don’t make resource identifiers dereferenceable. To write all those off “non-Semantic Web” really bothers me.

    Where’s the Web in these applications if there are no dereferenceable HTTP URIs, or they’re not somehow consuming, processing or exposing Linked Data? As I said above, that doesn’t make these applications wrong, in my view it just doesn’t make them Semantic Web applications. I could build a house out of car parts, but that doesn’t make it a car, it’s still just a house. Even if not a very nice one ;)

  7. John Goodwin

    I tend to agree with Tom’s last point (I think Heisenberg’s uncertainly principle applies here :))

    semantic * web >= h/2 (ok being a tad flippant)

    There are some great applications of OWL, or OWL and RDF out there – but the more OWL is used and the more semantic it is then the less “webby” it tends to be. Also IMHO linked data is not that semantic, but it is very web. Neither is write or wrong and both are great applications. I think with time is will be necessary to start seriously considering the place of OWL in the linked data web, and I would propose that any linked data web tutorials would encourage people to restrict themselves to either OWL 2 QL or OWL 2 RL (http://www.w3.org/TR/2008/WD-owl2-profiles-20081202/) based on the assumption that linked data will be about large volumes of data.

  8. Benjamin Heitmann

    I want to add another aspect to this discussion:

    If the term “Semantic Web” is becoming the next en vogue buzzword, then there is the big danger, that the term ends up having as little specificity as the term “web 2.0″ does.

    For the “World Wide Web” they layman could easily understand the requirements for something to be “on the Web”: if you can access it with a web browser, then it is part of the Web.

    How about “Web 2.0″? The layman might shortly wonder about the impact of the newness of the so-called second generation web. But then he quickly discovered that he can access it with exactly the same Web browser as the World Wide Web.

    Because the access mechanism for web 2.0 sites is exactly the same as for old school World Wide Web sites, people started adding whatever meaning they wanted to the term “web 2.0″, as it seemed like a non-technical distinction.

    And now the “Semantic Web/Web of Data/Linked Data” comes along. Marketers seem too like it, because selling something new makes the old stuff obsolete, which can drive a market.

    What about the layman? How can he distinguish the new Semantic Web from the old Web?

    The access mechanism does not help in distinguishing them: using a Web browser will yield plain old html pages, nothing new to see there. New and specialised ways to interact with the Semantic Web exist, but they are not marketed to the layman.

    So in the end the layman will see the Semantic Web as just another flavour of the Web, as a non-technical distinction.

    And at that moment in time, everybody will have the freedom to tack any kind of meaning to the term “Semantic Web”.

    The academic community might have been able to avert this, but if researchers cant event agree on the technical distinction between Semantic Web and non-Semantic Web applications, then it is unfair to propose any belated and non-inclusive definition of the term now.

    I guess, its to late for that discussion now.

  9. John Goodwin

    The bottom up and top down approaches to the “semantic web” discussed here:

    http://www.readwriteweb.com/archives/10_semantic_apps_to_watch.php

    Not sure I’d agree that the bottom up is necessarily academic or the top down was any more pragmatic though.

  10. Lee Feigenbaum

    “””
    Where’s the Web in these applications if there are no dereferenceable HTTP URIs, or they’re not somehow consuming, processing or exposing Linked Data? As I said above, that doesn’t make these applications wrong, in my view it just doesn’t make them Semantic Web applications. I could build a house out of car parts, but that doesn’t make it a car, it’s still just a house. Even if not a very nice one ;)
    “””

    To me, what makes something “Semantic Web” is simple: it is built on one or more of the W3C stack of Semantic Web technologies. I think people could argue until they’re blue in the face about whether an application is Semantic “enough” or Web “enough” to qualify for the label Semantic Web, but I also think that’s divisive and not productive. Non-dereferenceable, non-HTTP RDF triples can link together in a Web of Data just as well as HTTP URIs. It’s just another way of using the technologies, with its own merits and drawbacks.

    Apologies for the right/wrong characterization. I feel just as strongly if the discussion is limited to “qualifies for Semantic Web label or not” though.

    Lee

  11. John Goodwin

    Hi Lee,

    Hope my comment didn’t come across as arguing about whether things are “”Semantic “enough” or Web “enough”” – I more meant that (in my experience at least) applications based on RDF triples alone tended to be more webby that those based on OWL. Some semantic web apps are big on the semantics, some are big on web – not seen many examples of both. Where I work (OS in GB) we are looking at applications of OWL “behind the firewall” for data integration and modelling product specs/validation rules/feature catalogues. We’re also interesting in opportunities for OS and the linked data web. This work is more RDF and web based.

  12. Lee Feigenbaum

    John,

    Sorry for the confusion – I do agree with your observation, and tend to think that both ends of the spectrum can equally classify themselves as “Semantic Web applications.” I was replying to Tom’s comment. :-)

    Lee

  13. Ã…sa says » links for 2009-03-03

    [...] Linked Data? Web of Data? Semantic Web? WTF? – Tom Heath’s Displacement Activities (tags: attläsa semanticweb webben linkeddata) [...]

  14. Greg Boutin

    More on Ian Davis’s blog, and my views there.
    http://iandavis.com/blog/2009/03/the-semantic-web-acid-test/comment-page-1#comment-1362

    And on my blog here:
    http://www.semanticsincorporated.com/2009/02/the-siloed-view-of-the-semantic-web-as-linked-data.html

    Yes, I am at least partially a marketer (“boo oo, he’s involved with money… we programmers have pure intentions and don’t need to get paid…” Time to move on, this is a broken record…), although generally working on helping start-up with commercialization, and not running ads on soap opera. And yes, I think the attempt at restricting the semantic web brand are futile and it’s much better that way. Not b/c it can “make the old look new”, but because “semantic web” means a web that better gets the meaning of information, and that’s what people will look for. If that’s not what it does then rename it. But it’s too late for the semantic web, as others pointed out in this thread.

  15. Tom Heath

    Apologies all for the slow responses, been pretty tied up with other things. Will try to catch up now…

  16. Tom Heath

    Hi Greg,

    Picking up on your comments from 2nd March, I have to say I’m a bit confused. I’m surprised that, as a marketer, you’re so keen on using a term that has had a history as rocky as “Semantic Web”, and been subject to so much confusion and misunderstanding.

    You say in your most recent comments that “‘semantic web’ means a web that better gets the meaning of information”. I don’t disagree with this rather vague statement (though I still fail to see where the Web is in most of the apps you class as Semantic Web), but seriously, I’ve been explaining the Semantic Web to people for years and most just don’t ‘get’ this kind of explanation. It’s not the fault of those people, it’s just an abstract concept that is hard to grasp if you’re not immersed in the field.

    So a piece of advice from me to you: choose a term to explain your area of interest that makes sense to people, even if you use your preferred label when talking to peers. As I suggested above, something like “Smart Web” or “Intelligent Web” may be somewhere worth starting. These terms are far from perfect, but they’re at least a step closer to passing the elevator test.

    On the same note, you say that “the semantic web is a term that implies broader benefits than those linked data is capable of offering”. Seriously, again, I don’t think that for 99% of the world’s population the term “Semantic Web” implies any benefits whatsoever! The Semantic Web research community made this mistake already, and paid dearly for it. In contrast, talk to someone about using the Web to link data together and you’re likely to witness the kind of excitement and clarity of understanding that I’ve seen many times in the last few years.

    Switching focus to other themes in your comments… For the record, I have no problem with calling Linked Data apps “Linked Data apps”. But to purposefully shy away from also calling them Semantic Web applications would be ludicrous.

    I think you’ve missed a major point in my previous comments, so to try and put it a different way: I perceive that you see Linked Data and semantic technologies as somehow incompatible, perhaps even in competition for the hearts and minds of the world at large. What I said above, and will repeat here is that this kind of view is false; there is a pipeline or cycle that can feed the Web of data, which can in turn feed heaps of smart applications built on semantic technologies, which can then choose to re-feed the Web with more data. If these apps want to be part of the Web ecosystem then they need to have a way of being in the Web – not just delivered across the Web, but actually in it. This means publishing, or at least consuming Linked Data at some stage of the process.

    The Linked Data community is not trying to force-feed anyone with our technologies – we don’t have any technologies(!), only some guiding principles, a whole heap of enthusiasm, a bunch of successes under our belts, and a rapidly growing community of committed parties. The key technologies that underpin Linked Data are URIs and HTTP (these sound familiar?), RDF, and a little bit of RDFS/OWL.

    No one is forcing the groups that Lee and John describe to use any of these technologies, just as no one forced anyone to use URIs, HTTP and HTML the first time around. People chose to use these technologies because they enabled connections between, and access to, things that had never been easily connected or accessed before, and this brought huge benefits. I’m a strong believer in bringing these benefits to data as well as documents, and this is why I have strong views about what constitutes the Semantic Web. You don’t have to agree, but if you want your applications or services that are powered by semantic technologies to be silo-free and right there in the Web, then you’d best start investigating Linked Data interfaces for them. This is the best bet we have right now for building a Semantic Web.

    P.S. Yes, please can this false dichotomy between marketers and programmers please be put to rest once and for all. It’s not a helpful or accurate way of characterising the world.

  17. Tom Heath

    Hi John,

    Thanks for your comments and pragmatic view on the topic. I look forward to your input in thinking through how the Linked Data community needs to respond to and can make use of OWL2. I think I owe you an email reply on that, right? ;)

    Regarding the top-down bottom-up distinction, I (being brutally honest here) think it’s an utterly worthless model for understanding, characterising or communicating about the Semantic Web (or whatever one wants to call it). Some of my issues with it are listed here: http://www.readwriteweb.com/archives/rdf_semantic_web_apps.php#comment-112737

    Cheers, Tom.

  18. Tom Heath

    Lee,

    I guess we’ll have to agree to disagree (though not by as much as I disagree with Greg ;) Hehe. Just as one final question though, from a technical rather than privacy/security/intellectual property perspective, why would any of these apps built on the SW technology stack not want to be in the Web, Linked Data-style?

    Feel free to save the answer for over a beer in October if you feel like you’ve expended enough energy on this already :)

    Cheers, Tom.

  19. Greg Boutin

    Tom, I am sorry you feel like disagreeing so much with me. Disagreeing with Greg Boutin seems to be a fashionable theme in the linked data community these days, and certainly the fact that you repeat it again proudly in the last comment doesn’t make me feel like you’ve taken an objective stance on what I actually said. Having said that, if you’re open to potentially agreeing on something, there is good news b/c you seem to disagree on your own conclusions about what I said, and not really about what I said.
    1. “as a marketer [note I've said I was only partially one...], you’re so keen on using a term that has had a history as rocky as “Semantic Web”” Where did I write that again? I’m not finding anything saying that in my comments, which pertained to linked data not having exclusivity on the semantic web brand.
    2. “I’ve been explaining the Semantic Web to people for years and most just don’t ‘get’ this kind of explanation. It’s not the fault of those people, it’s just an abstract concept that is hard to grasp if you’re not immersed in the field.” Well, again, that has nothing to do with what I was commenting on. Having said that, I explained the semantic web to many folks too in these terms, and provided you add a couple examples they generally get it. So with all due respect the problem may not be that it’s “hard to grasp if you’re not immersed in the field”, but that it’s “hard to explain if you are immersed in the field”. That’s a well-known problem I blogged about (search for The Knowledge Curse on my blog, and read Made to Stick)
    3. “On the same note, you say that “the semantic web is a term that implies broader benefits than those linked data is capable of offering”. Seriously, again, I don’t think that for 99% of the world’s population the term “Semantic Web” implies any benefits whatsoever!”
    Again, all peripheral facts, but it’s ok to discuss. The question is what would people think when seeing “semantic web”. Semantic, which is meaning, and web, which is the internet. That’s what people who understand those words will get. So, it may not be the best way to brand it, but it certainly conveys benefits of a technology that shows stuff on the web based on deeper meaning. The fact no ones knows about it today, beside being increasingly wrong, has nothing to do with it.
    4. “I perceive that you see Linked Data and semantic technologies as somehow incompatible, perhaps even in competition for the hearts and minds of the world at large.”
    You assume things I did not say, write or believe. I am sorry to see you invent things like that about what I wrote, just like others in the community have in some rhetorical attempts to associate me with such silly ideas. I don’t think in any way that linked data and SW are incompatible. I do see Linked Data as the most positive thing that happened to the SW and the most important part of it at present. I just don’t see Linked Data as a technology that should be imposed upon everyone (note I’m not saying you personally advocate for that), and I see a lot of SW apps being SW without using linked data. I just disagree with you that being SW means “publishing, or at least consuming Linked Data at some stage of the process.”
    5. “The Linked Data community is not trying to force-feed anyone with our technologies.” I didn’t say it was. I said trying to push the linked data is a must for SW idea could “be seen as an attempt to force-feed the technology”
    6. “You don’t have to agree, but if you want your applications or services that are powered by semantic technologies to be silo-free and right there in the Web, then you’d best start investigating Linked Data interfaces for them. This is the best bet we have right now for building a Semantic Web.” I don’t dispute that Linked Data may be the best way, see my comment to Ian, but I think linked data should not be forced top-down, and an app may still claim to be SW without using it, which you disagree with. What you had written earlier was that “Anything claiming the Semantic Web label needs to get its hands dirty with Linked Data somewhere along the way. That’s just how it is.”

    Overall, I hope it’s clear we simply disagree on what can use the term SW and what can’t. And I hope you won’t join the ranks of those 2 or 3 other linked data folks who seem to have decided that whatever I asserted they’d be opposed to it… out of principle (or lack thereof). Thanks for the discussion Tom.

  20. John Goodwin

    Hi Tom,

    Pretty much agree 100% with you comment:

    http://www.readwriteweb.com/archives/rdf_semantic_web_apps.php#comment-112737

    In fact some of it could have been written by me :) I often mention at work that the SW is pretty much at the same stage the web was back in the early/mid 90s. For me the linked data cloud is pretty much analagous with those directories of the web you could purchasse from all good book shops.

    Given the definition of the SW I personally adhere to and understand a lot of current, so called, SW applications are not really SW. This does detract from them being good applications, just not SW applications.

    Still the beauty of ontologies is we can agree to disgree. Shall I construct an OWL ontology defining a SW app? :)

    I would, however, love to see more applications of RDF in a linked data stylee. There are a lot of us putting data up, there are ppl like you guys at Talis, Sindice, Zitgist and Virtuoso desiging some great tools/platforms…but it would probably be fair to say there are thus far no REALLy great RDF applications out there on the web. Fair comment?

  21. Tom Heath

    Hi Greg,

    I’m not disagreeing with you to be difficult, and certainly not to be fashionable. I’ve read what you said, and so far it hasn’t changed my views. Your comments imply that if I was approaching this objectively I would have done so, but surely the same would apply to you with respect to my comments. We just happen to differ in opinion, and that’s fine, a natural part of the discourse.

    Re 1. You didn’t write this, it’s an impression I’ve formed, that you’re keen to align yourself with the “Semantic Web” label. Taking the historical view I’m curious about that, compared to other potential labels.

    Re 2. Clearly I can’t judge this objectively, but my subjective experience tells me that starting from the term “Linked Data” offers less resistance to understanding than starting from “Semantic Web”.

    Re 3. In my experience people find it hard to derive the desired meaning from the term Semantic Web, without receiving further explanation, even if they do break it down into constituent parts. If people are still conflating the Web and the Internet I do wonder what they make of “Semantic Web”, but if you’ve had more success that’s great.

    Re 4. Greg, all I said was that I perceived something. If you don’t see the two as incompatible then that’s great – we agree on something at last :)

    Re 5. A useful clarification, thanks.

    I don’t think there’s anything to add regarding 6. Let’s agree to disagree and get on with building this Semantic Web and its associated applications, whatever definitions we choose. I might continue to disagree with you, but please rest assured it would be on the merits of the arguments rather than on any principle of disagreement. Thank you too for the discussion :)

    Tom.

  22. Alan Morrison

    Tom, I generally agree with you on this, but we’re writing an issue on enterprise use of semantic Web standards, and it occurs to us that Linked Data is a lousy term from an SEO perspective. So is “data Web,” for that matter. Linked Data alludes to something specific using words that couldn’t be more vague.I find it necessary to use all three terms and augment them with explanations and examples to get points across to various audiences. The lay audience doesn’t know to search on “linked data” rather than “semantic Web.”

    Moreover, my feed reader has both the SW and LD terms, and I’ve discovered that I miss a lot of relevant info if I just use LD and exclude SW.

    “Semantic Web” doesn’t describe what it is, but how it does it, and how it does it is unfamiliar territory to most, including developers. That’s why it’s inscrutable from a layperson’s perspective. Describing the migration from document to data Web at least has a visual association.

    BTW, this is a typical dilemma for a software subject, particularly such an ephemeral one. We do the best we can with examples and metaphors, and hope some people have the patience to ponder what we say.

  23. Interesting stuff from around the web 2009-03-20 « Derivadow.com

    [...] Linked Data? Web of Data? Semantic Web? WTF? [Tom Heath] “Think about HTML documents; when people started weaving these together with hyperlinks we got a Web of documents. Now think about data. When people started weaving individual bits of data together with RDF triples (that expressed the relationship between these bits of data) we saw the emergence of a Web of data. Linked Data is no more complex than this – connecting related data across the Web using URIs, HTTP and RDF.” [...]

  24. 시맨틱 웹, 그리고 링크드 데이터 « Passion for Information

    [...] Issues: Linked Data“, 2006, http://www.w3.org/DesignIssues/LinkedData.html 2. Tom Heath, “Linked Data? Web of Data? Semantic Web? WTF?” http://tomheath.com/blog/2009/03/linked-data-web-of-data-semantic-web-wtf/ 3. Passant, [...]

  25. Internet Alchemy » The Semantic Web Acid Test

    [...] Heath writes a cracking post on the current attempts by a few people to brand web applications that happen to perform text [...]

  26. The Semantic Web Acid Test | Internet Alchemy

    [...] Heath writes a cracking post on the current attempts by a few people to brand web applications that happen to perform text [...]

  27. FAQ on Linked Data – Connected Distributed Data Across the Web | The Book

    [...] Linked Data? Web of Data? Semantic Web? WTF? – by Tom Heath [...]

  28. dipendra singh

    Hey Dear Tom,
    m doing M.tech thesis from DA-IICT india on linked data.So may please give me very basics of what is linked data as m new in this area.thank u

  29. Tom Heath

    Have a look at my Linked Data Book.