Traditional databases are highly structured and centrally controlled. You can type things into existing forms but only the IT department can add new types of records. The semantic web takes a different approach. It provides a vocabulary to describe the information and relationships stored in databases. Data owners can use this vocabulary to label their information and publish it online. As the description and vocabulary are public, any program can process the data; just as any web browser can display a website. Similarly, information from different sites can be merged and processed together.

Information in the semantic web is described using RDF (Resource Description Framework). It is a language that lets people describe data. For example, it might provide a bibliographic description of this article: author, publication, date and title. HP Labs has co-developed SPARQL[i] (SPARQL Protocol and RDF Query Language, pronounced ‘sparkle’). SPARQL is the query language and access protocol of the Semantic Web. HP has lead the development of the SPARQL specification. It is a language that programmers can use to search RDF descriptions in web pages, as if they were records in a more traditional database. “SPARQL will make a huge difference,” said Berners-Lee, in an interview for ZDNet.co.uk.

Everything in this series of articles exists today, albeit in prototype form. This is the job of a research lab – to make the future visible. IF HP Labs are right, their research – virtualisation, trusted computing, utility computing and the semantic web – offer a vision of IT that is substantially different from what we see today. Computers will process data on a global scale; they will get more done, more efficiently. They will be reliable and trustworthy, and you will be able to rent supercomputer performance by the hour. William Gibson said that the future is already here, it’s just unevenly distributed. If you’re looking for it, try HP Labs Bristol.