Arthur Cole spoke with Dr. Mark Butler, senior lecturer in Information Architecture at the University of the West of England.
Cole: There's been a lot of talk about the semantic Web for a number of years, but may people are still in the dark. How would it change the Web as we know it today?
Butler: In my view, the change is here already, because I would define the semantic Web as people achieving competitive advantage through others re-using their data. A good example here is Amazon: They have relentlessly improved their book database by adding information like cover images, reviews etc. Now lots of other Web sites and programs, for example, LibraryThing or Delicious Library, re-use that information. Tim O'Reilly, in his paper on Web 2.0, identifies exactly this kind of data re-use as a key part of Web 2.0.
Of course, the semantic Web purists would not agree because people like Amazon are not using semantic Web standards, and they say those standards are essential because they allow computers to understand "meaning" in some magical way. I think that is nonsense - what we are dealing with here is data, and the main point about the standards is they should simplify the kind of re-use we are already seeing.
Cole: Are there any benefits to commercial enterprises and e-commerce?
Butler: For the semantic Web standards, yes. When XML came out, it was billed as a way of solving enterprise integration problems. However, it was not quite true: When you exchange XML documents, you still need to agree on the vocabulary and the structure of the document. In the semantic Web equivalent, RDF, you only need to agree on the vocabulary. This makes it easier to exchange data, combine it and automatically generate code to process it, all simplifying the problem of enterprise integration.
The reason we do not see RDF more widely used at the moment is a bit like a re-run of VHS versus Betamax. Even though RDF (Betamax) is technically better, it is not quite compatible with XML (VHS), and XML is so much more widely used that it has been hard for RDF to gain a foothold. But it is used: It is surprising the number of companies that are already investigating its use in one project or another.
Cole: What do you see as the next step in development of the semantic Web?
Butler: I think there are three things that are really important at the moment: one is wider use of the N3 serialization of RDF. This is a much better, more human-readable way of writing semantic Web data. The second is better tools for working with RDF to create, validate and browse data, because these tools form re-usable building blocks of creating semantic Web systems. The third is efficient RDF stores. The research community has been making good progress on this, coming up with ways of storing RDF efficiently in relational databases, and converting queries written in semantic Web query languages like SPARQL to SQL.
Collectively, these developments will mean we start to see more semantic Web systems built using semantic Web technology. It will also mean those systems were easier to build, and are enterprise scale, not just small-scale prototypes, which in turn will encourage further uptake.