Web 2.0 + Web 3.0 = Web 5.0: SADI & CardioSHARE: A New Perspective on Semantic Web Services

Mark Wilkinson



The "Deep Web" is a term used to refer to all of the data on the Internet that is hidden behind interfaces and/or generated by analytical algorithms, and is therefore inaccessible to search engines. The quantity of data in the Deep Web is estimated to be several orders of magnitude greater than that on the surface Web, and it is therefore critical that we carefully consider how to make these resources available as we attempt to build the Semantic Web for healthcare and life sciences. Our experiences running the BioMoby project have taught us a great deal about the nature and behaviours of Semantic Web Services in the bioinformatics space.

We have now applied this experience to a new Semantic Web Service framework - SADI (Semantic Automated Discovery and Integration) - which attempts to expose Deep Web data in a straightforward, standards-compliant manner. SADI does not define any new technologies, message structures or ontologies; it merely circumscribes a set of recommended practices for bioinformatics service providers that result in highly interoperable behaviours between disparate resources on the Semantic Web.

In this talk I will present the rationale behind the SADI middleware, and show its interoperable behaviours through demonstrations of two very different client applications. Finally, I will introduce the CardioSHARE project, where OWL-DL ontologies are used to drive the discovery of cardiovascular data and analytical tools on the Semantic Web via the SADI framework. CardioSHARE demonstrates the significant additional power obtained through defining ontological classes based on their properties rather than asserted IS-A hierarchies; we hope that such demonstrations will provide incentive for the community to move toward these richer ontological descriptions.