I have worked in the IT industry since 1996 as a developer, architect and consultant software architect. I have provided consulting and auditing services to clients across various industries (financial, healthcare, mining, publishing etc) ranging from detailed code analysis to high-level conceptual architectures.
Currently I am the Ontology Tools Technical Lead at the European Bioinformatics Institute (EMBL-EBI).
You can visit my Linkedin profile for further information about me.
My Aim with this Blog
I strongly believe that the Semantic Web is the way to unlock the wealth of information that is currently locked up in isolated websites. Unlocking this information will take society forward in ways that we cannot imagine. However, to realize the Semantic Web we need to craft the web of data, which requires large scale adoption of Semantic Web Technologies by software developers. The problem is that the underlying theory and plethora of Semantic Web tools present software developers with a steep learning curve.
After 10 years involved in the Semantic Web in one way or the other I still have a lot to learn about Semantic Web Technologies. From my volunteer work at Stack Overflow I realize that I am not alone in this regard. It is my hope through this blog that I can learn from and help other Semantic Web practitioners.
I have a Ph.D in Artificial Intelligence/Data Science. My M.Sc dissertation and Ph.D thesis can be found under downloads.
In my M.Sc dissertation I investigated how Semantic Technologies, and in specific Description Logics, can be used to validate UML class diagrams and detect modeling heuristic violations.
My Ph.D topic deals with generating/extracting “perfect test data”. The idea of “perfect test data” is mathematically precise in that it will ensure that the test data satisfy all constraints that hold and violate all constraints that do not hold for a given class of constraints. In the case of uniqueness constraints for RDF datasets this can result in an example dataset that is 0.00003% of the original dataset while preserving the constraints of the original dataset perfectly.
My motto: “Ignoring theory has a tendency to fail in very practical ways”.