The Semantic Web : One Mo’ Time!

There’s a bit of buzz today around a new startup called Metaweb. Metaweb is creating a new type of semantically-flavored version of Wikipedia called Freebase (heh).
In a nutshell, Freebase allows the community to contribute to facts about stuff, like Wikipedia, except with one key difference: the system wants the stuff in some semantically connected way.
We’ve actually chatted in the office here about how Wikipedia comes oh-so-close yet it’s a near miss. There’s all that great stuff yet, beyond a few hyperlinks, none if it is semantically connected. Tim Berners-Lee, has been introducing (and re-introducing, and re-introducing) the “next Web” or the Semantic Web for years now. And it still hasn’t caught on. The eureka for me was FTrain’s excellent semi-sci-fi scenario of how Google crushed everybody by capturing the semantic web.
For me, here’s the thing with the semantic web: it’s like Communism. It’s really great on paper. But in practice, it’s rife with problems half-filled promises. I think the crux of it, ironically, is usability. For something to take off, it must provide some sort of immediate, conceivable value to the masses upon touching it. Yeh, it can sound really cool in a white paper, but guess what? Nobody reads white papers. People are willing to do stuff that makes sense to them in an immediate way.
The most popular RSS reader in the world isn’t an RSS reader and makes no mention of being one: My Yahoo! collects articles and puts them in one place. It’s convenient and can easily be understood. Who cares about the inner workings of RSS?
For the semantic web to become real, we need to get the data in first. And that’s a somewhat daunting task. All the evangelizing in the world won’t galvanize people to get started. Instead, tools like Freebase and Google Base bring it down to earth for the average user. Recipes. Used cars. Stuff that may be worthwhile. Why else would be people contribute to the semantic web?
And there lies the rub: if you look at Google Base today, it’s just plain…weird. It’s essentially a collection of smaller database-style apps that don’t seem very connected at all. Can I be perusing their personals section and find a recipe from one of the dating candidates? Or a used car they’re trying to sell? Maybe the better question is: do I really want to?
For the general population, it feels kind of, well, half-baked. How do we get the masses to pitch in not only on the “data entry” side but on the creation of the semantic structure as a whole? Considering the usability challenges of making the semantic web take off as it is, maybe this is too lofty a goal.
Either way, like RSS, the barrier to adoption isn’t technology or the capabilities of systems (though there are some interesting puzzles). The barrier is people. Like Berners-Lee’s other invention, HTML, we need a killer app. For the masses, the browser was the invention, even though an underlying hyperlinked web was the real magic. The Semantic Web is in need of its own killer app.

4 Comments The Semantic Web : One Mo’ Time!

  1. Aran

    I’m a reasonably smart person who develops websites and has even created some moderately complex relational databases. I’m even left-handed, which means I’m generally comfortable with non-linear methods of organizing data.
    Yet, I’m more or less scratching my head over what the semantic web is. At this stage it is just WAY too abstract a concept for this to be anything more than a fun toy for engineers to fantasize about.

    Reply
  2. Tom Morris

    Aran, that’s what I’m trying to solve with getsemantic.com – a new project to help web designers and developers to grok the semantic web and implement stuff.
    I don’t think that companies like Freebase are going to be a big influence in building the SemWeb – more likely, the technology will get simple enough and understandable enough for individuals to just start publishing it themselves on blogs, wikis and old-fashioned web pages. Microformats are one way that’s going to happen. Domain-specific providers are another way. GRDDL is another. SPARQL endpoints is another.

    Reply
  3. PTC

    Semantic web is a great vision that can bring revolutionary benefits to internet users. It certainly lacks on immediate benefits to its followers; but considering adapataion happening from both – Academia and Industry, it is definately coming. When it is coming is not known yet though.
    A comprehensive take on Semantic web and Web 3.0 at http://techbiz.blog.com/1730241/

    Reply

Leave a Reply to Tom Morris Cancel reply

Your email address will not be published. Required fields are marked *