Workshop on Challenges and Promise of Semantic Web Services

Yesterday I attended a one day workshop on semantic web services.  It was interesting to take part in discussions in an academic context.  One of the things that struck me was the desire of the researchers present to make sure that the work they were doing had relevance to the "real world", by which they meant commercial applications.

I started off the day and was followed by Dr Monika Solanki from De Montford University.  Monika spoke about the challenges facing composition of web services in general and use of semantic web technologies in web services composition in general.  At the end of the presentation you could either be very depressed by all the problems she identified, or very excited by all the research opportunities.

An interesting question discussed was why have web services taken off in the commercial world but not semantic web.  We didn't quite generate the passion of the recent Tim Berners-Lee/Peter Norvig 'debate'.  Some suggestions we had were

• Web Services fit into an existing distributed computing model, making them easier to understand.  The "What is an ontology anyway?" argument


• Web Service tools (like Visual Studio and JDeveloper) hide much of the complexity of WSDL/SOAP and BPEL.


• Web Service standards are coalescing rapidly.


• What problems are semantic web technologies good for - dynamic service composition raises the question of would you trust the generated composite service - if your answer is yes would it change if you had to pay for the services the composite consumed?

I thought one of the most interesting comments was that "maybe Semantic Web needs marketing".  To which I believe the answer is possibly yes, but we need to focus on real problems that affect people now and it may be that widespread semantic web acceptance is still 3-5 years away.

Oracle has an interest in semantic web, particularly in the pharma space.  Take a look at the Oracle Semantic Web pages.  This site also has links to information on support for semantic web technologies (including indexing) inside the Oracle database.  So maybe there is some commercial interest after all.