Reflecting On Openness and the Semantic Web
Posted by Brian Kelly on 12 Apr 2008
The printed copy of the proceedings of the Museums and the Web 2008 conference divides the papers into four sections: Institutions, User Participation, Web Space and Reflecting. The concluding section, on Reflecting, contains only two papers: one on Semantic Dissonance: Do We Need (And Do We Understand) The Semantics Web? by Ross Parry (University of Leicester), Nick Poole (The Collections Trust) and Jon Pratty (Culture 24) and my paper on What Does Openness Mean To The Museum Community?, co-authored by Mike Ellis (Eduserv) and Ross Gardler (JISC OSS Watch), which I’ve posted about recently.
It is pleasing that the two papers which reflect on the challenges and opportunities posed by recent Web developments have been written by a combination of researchers and practitioners based in the UK.
Ross Parry’s paper is based on a series of workshops funded by the AHRC which were held at various locations in the UK during 2006 and 2007. The paper describes discussions which have taken place recently in the UK in which it has been suggested that “museum data with good URIs, consistent metadata and simple tagging are seen to provide a vitally stable infrastructure on which to build“.
To this list I would add the importance of providing data which is free from restrictive licence conditions and which is exposed for reuse by other applications which can exploit the rich semantic data.
But stable URIs, consistent metadata, simple tagging, open data and machine interfaces – isn’t this what Web 2.0 is about? From one perspective, people may regard Web 2.0 as shorthand for referring to blog, wiki and RSS applications. But Tim O’Reilly’s original Web 2.0 diagram makes it clear that Web 2.0 is broader than this.
In a chapter entitled ‘‘If it quacks like a duck…’ – developments in search technologies‘ in a recent Becta Research Report on Emerging Technologies for Learning Volume 3 (2008) (PDF version of chapter) my colleague Emma Tonkin argues that:
By “semantic”, Berners-Lee means nothing more than “machine processable”. The choice of nomenclature is a primary cause of confusion on both sides of the debate. It is unfortunate that the effort was not named “the machine processable web” instead.
I think Emma is right: the term Semantic Web has caused much confusion. But if the Semantic Web is really a machine processable Web in which clean URIs can help to provide programatic access to structured data, then isn’t this very close to what Web 2.0 may be considered to be about?
And can you claim to be in favour of the Semantic Web if you are critical of the architectural aspects of Web 2.0? Or, to put it another way, isn’t engagement with Web 2.0 a needed stepping stone towards the Semantic Web? And won’t we find that those who come out with reasons for not engaging with Web 2.0, will come out with a similar set of reasons for not engaging with the Semantic Web?
PeteJ said
I don’t really understand what you mean by the “the architectural aspects of Web 2.0” – I’m not sure there are any unifying architectural principles of “Web 2.0”! – so I don’t know whether I’m critical of them or not. :-)
I do know that some of the data I get from some apps which describe themselves as “Web 2.0” isn’t as readily machine processable as most of the data I get from apps which describe themselves as “Semantic Web” apps, so at the “technical” level there is certainly a difference.
If you are saying that many of the “social”/”cultural” principles (openness, etc) which are associated with/necessary for “Web 2.0” are also necessary for the “Semantic Web” then you may be right.
And I’d certainly agree that some technologies and applications which describe themselves as “Web 2.0” are also (what I think of as) “Semantic Web” technologies and applications.
But do I think “Web 2.0” is the same thing as “Semantic Web”? No, TBH, I don’t.
Brian Kelly (UK Web Focus) said
Hi Pete – I was referring to characteristics such as ‘cool URIs’ and use of clean HTML and CSS, unlike the application-specific URIs and presentational HTML which characterised Web 1.0.
PeteJ said
Ah, OK. Well, I agree that “Cool URIs” are pretty much a sine qua non for the Web, whether it’s “2.0”, “3.0”, “3G” or “Semantic”.
And while I think CSS is a wonderful thing, it seems to me the use of CSS is somewhat orthogonal to the question of the frameworks and formalisms we choose for representing data on the Web.
Owen Stephens said
I would guess that most of those involved in discussions about the pros and cons of the semantic web believe that we need to have a ‘machine processable’ web. The question is what kind of processing you want to be able to do, and how you achieve this.
Many of the arguments seem to focus around an ‘idealism’ vs ‘pragmatism’ divide – those that think we should all be publishing web pages with nice semantic markup using XML and RDF, and others who believe this will never happen (for several reasons) and we need to focus on more realistic goals.
I guess at the one end of the spectrum you have those who believe all publishers on the web should publish in a standardised/structured way that is machine processable now, and at the other end those who believe that publishers should publish whatever they like and we have to get the machines to process it. It’s interesting to see that Yahoo seems to be positioning itself on the structured publishing side and Google firmly on the better machine processing side.
Andy Powell said
It strikes me as rather unhelpful to try and suggest that ‘semantic’ could/should have been ‘machine processable’. ‘Machine processable Web’? What does that mean? The Web in all it’s forms has been ‘machine processable’ – I don’t want to shatter any illusions but Google indexes aren’t hand-crafted by people! :-)
Brian Kelly (UK Web Focus) said
Hi Andy – The mini-workshop on “Semantic Dissonance: Do We Need (And Do We Understand) The Semantic Web?” facilitated by Ross Parry at MW2008 began by trying to seek consensus on what was meant by the term ‘Semantic Web’. The ‘machine-processable Web’ was one of the suggestions made – but no consensus was reached.