Twitter Annotations & The Semantic Web

A few months back, I was on a call with other Semantic Web meetup organizers from around the world. We were all discussing the perceived lull that’s happening with the Semantic Web. The general theme was that people were talking about it less and coming up with less visible applications. Interest was waning and there was a fear that the SW is far back in the minds of non-academics.

Well here is a great example about just how far back the semantic web is: Twitter just announced an ‘annotations’ feature in their upcoming API. From their group:

“First off let’s be clearer about what an annotation is. An annotation is a namespace, key, value triple. A tweet can have one or more annotations. Namespaces can have one or more key/value pairs. ”

Now, on its face, that sounds like good news for the semantic web. However, read down a little bit and you’ll see that they’re not talking about RDF representations:

We’re thinking we’ll provide two mechanisms for specifying what a tweet’s annotations are:

  1. JSON
  2. form encoded parameters

I can’t think of better use case for the semantic web in general and RDF in particular than these annotations. They even used the term ‘triple’ to describe it. RDFa would be be a great embedded metadata format for presenting these annotations on their website. But to not even mention the RDF in the first wave of implementations shows just how far back the semantic web is in the minds of one of the most visible web sites in the world.

Now, they don’t dismiss RDF & the SW entirely:

“This isn’t final. The payloads could end up wildly different after we noodle around in things like RDF and the semantic web’s literature and all that kind of stuff. You can’t see me but my hands are waving vigorously. ”

However, that sounds pretty non-committal to me. I can easily see a situation where they come back saying that RDF & the SW is too complex, heavyweight or confusing to roll out as the primary representation in their API. If Twitter annotations succeed at providing widespread, structured metadata without formal support for RDF / the SW, could that start the of the decline of the Semantic Web for mainstream usage?