27 Oct 2009
A new Internet language will enable computers to unravel the complex terminology used by experts such as doctors, engineers and life scientists, and understand what they really mean.
OWL 2, which was developed by an international team led by computer scientists from the University of Manchester and Oxford University, is a language like the HTML that makes up most of the web pages we use.
But OWL 2 is different because it is designed to enable computers to understand and interpret the contents of its pages rather than just display them for the benefit of humans.
One of the first and most important applications for the new language is in helping computers to understand and analyse specialised medical terms.
"The World Wide Web as we see it today is rather like a collection of linked documents,’ said Professor Ian Horrocks of Oxford University’s Computing Laboratory who helped to develop OWL 2.
"Whilst humans are very good at analysing the data contained in these pages, languages such as HTML do not help computers to ‘bridge the meaning gap’, and understand that, for instance, ‘paracetamol’, ‘acetaminophen’, and ‘para-acetylaminophen’ are all names for the same thing."
A good example of the scale of the problem facing medics and computer scientists is the NCI Cancer Thesaurus that has swollen from 20,000 medical terms in 2004 to over 50,000 terms today.
As new terms are being added all the time, ensuring that all these terms are described, updated and linked together correctly is a mammoth task for humans. However, by using OWL 2 definitions can be written in such a way that computer programs can tirelessly update these terms, enriching the structure of the Thesaurus and pointing out where there are errors.
"The first stage was writing the NCI Thesaurus in the original version of the language, OWL, but now OWL 2 enables computer programs to interpret these terms in a much more human-like way, for instance reasoning that if a fracture is located on a bone which is part of a leg then that fracture is a fracture of that leg," said Bijan Parsia of the University of Manchester’s School of Computer Science.
"It may sound simple to make this connection to us, but it’s a great leap forward that computer programs will be able to reason and make connections in this way.’
The development of OWL 2 is part of continuing efforts by computer scientists to build the ‘Semantic Web’, a Web that suits the needs of computer programs rather than just the needs of human web users.
OWL 2 is set to reach the final stage of ratification by the Web’s international standards organisation, the World Wide Web Consortium (W3C), this week.
This ‘Recommendation’ stage means that the standard has undergone extensive review and testing and is now endorsed by the W3C as a standard suitable for widespread use.
Notes for editors
For further information contact Professor Ian Horrocks by email ian.horrocks@comlab.ox.ac.uk or [mobile] +44 (0)783 7386916 NOTE: Professor Horrocks is in the USA this week so an initial email is best
Alternatively, contact the University of Oxford Press Office on +44 (0)1865 283877 or email press.office@admin.ox.ac.uk
Or contact the University of Manchester Press Office on +44 (0)161 2758387 or email alex.waddington@manchester.ac.uk
The first version of OWL (Web Ontology Language) was standardised by the World Wide Web Consortium (W3C) in 2004, see: http://en.wikipedia.org/wiki/Web_Ontology_Language
For a full list of all those involved in OWL 2’s development go to http://www.w3.org/2007/OWL/wiki/Participants
SOURCE