The Semantic Web
A decade after proposing the World Wide Web as a fun little project (and a scant few years after it went mainstream), Tim Berners-Lee was back up on his proposing horse again with his next jolly wheeze: the Semantic Web. As with the original, the ideas behind the Semantic Web were simple.
The overarching idea was that, just as the current Web was envisaged as a web of documents, the Semantic Web would be a web of data. All that it requires is for data to be expressed as Resource Desription Framework (RDF) triples; a chain of subject → predicate → object terms, for example:
This article →was authored by → John.
The predicates (properties or relationships) linking subject and object can be expressed unidirectionally:
John → is the author of →this article.
And you can daisy-chain RDF triples
This article →is about →unemployment.
Unemployment →is the subject of →these articles.
[… and so on.]
Now you can't express everything that is important to know about the world in this way (poets, painters, and musicians will not lose their jobs because of the Semantic Web — in the unlikely event they currently have paid jobs in these vocations), but you can do a lot of useful things.
Moreover if:
- the possible kinds of entities and the relationships between them are defined by one or more standardised vocabularies,
- the data is expressed (serialised) in one (or more) machine-readable formats, preferably non-proprietary, open standards
- individual entities are given a universally unique identifier (UUID), and
- this UUID is directly expressed as (or can be dereferenced to) a Web URI
… you are well and truly off to the races with a Semantic Web of linked data. The whole Web is potentially one unimaginably big database, and you can have your own software on your own computer using that big database (more or less) as readily as if it were a little database on your own hard drive. (I am oversimplifying; but that's the dream, at least.)
The Semantic Web was envisaged, just as was the original Web, and the Internet upon which it was built, as a dumb assemblage of network hardware, standardised protocols, languages, and so on. All the system was supposed to do was facilitate the sending and receiving of resources between end users. It was the users of the system that defined what the system did, because it was the users who built, ran, owned, and controlled the value in the system, not some central administration. (In fact, I recall TimBL saying that he intended that Web server software would come as a package with Web browser software; he expected that it should be as technically trivial to run a web site as it was to read one.) The Web was a "World of Ends".
However, it turns out that some parties had a different view of the ideal network architecture.
"Web 2.0", Software as a Service, and the Cloud
