Web 3.0 – Some useful distinctions and what is an ontology, anyway?

I made a short film on the story of the Semantic Web because I was fascinated by the philosophy it was based on and by the people who devoted themselves to it. We think of technology as something “other” than us, artificial as distinguished from natural, so I hoped to call attention to the fact that technology is built by people. John Hebeler’s understanding of the Semantic Web as being “all about relationships” and Clay Shirky’s objection to it because he doesn’t think we can “unambiguously describe the world” illustrate my belief that technology is less an exact science than the expression of a worldview.

But of course it’s more complicated than that. To tell a better story, I chose to leave a lot out (I can assure you that you wouldn’t have wanted to watch the hour-long more “nuanced” versions I started out with). I’m going into some of that here.

I want to complicate the objections raised by “the critics” of the Semantic Web. Essentially, the argument I presented was that while getting all the information on the web into standard formats might be a great idea in theory, in practice it can’t work because people don’t agree about the definitions of anything. In fact, the Semantic Web doesn’t require everyone to agree on the same ontologies. (In case you didn’t catch what an ontology is, think of it as a taxonomy that’s more specific about the relationships between things.) The idea is for everyone to build up their own ontologies, but to have some standard ways of connecting them together (so I can say the “Kate Ray” on Facebook and the “Kate Ray” on Twitter are the same person, and be able to bring together information from different sites).

Having lots of ontologies still doesn’t resolve everything. It actually gets into an even headier controversy about what it means for one thing to be the “SameAs” another thing (see my post on ‘Ontology Alignment’), but it’s not indicative of a failure to acknowledge that people have different worldviews. The “neatest” of the Semantic Web academics I talked to (including Frank Van Harmelon, whose interview didn’t end up in the film) may have leaned toward a more scientific, objective view of the world, but they weren’t suggesting that we eliminate diversity of thought. Even so, there are plenty of areas where people should agree about definitions. A Japanese factory working with an American factory to build airplanes should probably have some pretty damn specific agreements about what each part is and how it fits in with the other parts. Manufacturing and other enterprise-level areas, in fact, are currently hot applications of the Semantic Web.

An earlier version of the film also distinguished between the Semantic Web vs Linked Open Data. The Semantic Web is closer to the original vision – all information rendered machine-understandable so that computers can reason across huge swaths of data and potentially discover things we can’t now. Linked Data is a more toned-down version or the first step – just get information free from the applications that are keeping it locked up, and figure out what to do with it later. Linked Data is actually what Tim Berners-Lee was talking about in the TED clip I showed and in the year since his talk, quite a lot of organizations have added their data, from The New York Times to the U.S. Government. Whatever the debate surrounding the usefulness of the Semantic Web as a whole, the response to the Linked Data project clearly seems to be a step toward better transparency.

I could continue on this post in ever-refining detail (and an ever-diminishing audience), but instead will just repeat that my motivation for all this is discussion. Technology has important implications for how we think about the world, which is why I think these are the kinds of conversations we should be having. Thanks to everyone who commented, wrote posts in response, and sent me emails. I intend to keep up my side of the dialogue.

Advertisement