Ontologies are a structured way of representing knowledge. Ontologies are being developed and exposed on the Web to support a variety of applications, including biological knowledge sharing, enhanced search and discovery, and rapid enterprise integration. Ontology alignment is the process of discovering relationships between elements across two or more ontologies based upon a variety of techniques that use entity labels, structure, semantics, usage and external resources.
Suzette Stoutenburg, who graduated with a Ph.D. in Computer Science in May 2009, developed algorithms to align ontologies that go beyond traditional similarity and equivalence relationship. These algorithms use machine learning techniques such as support vector machines. She also experimented with scaling her algorithms to work with large-sized ontologies. We have published several papers based on her work. Here is an example paper: International Journal of Bioinformatics Research and Applications 2010 paper.
Our future work will involve working on alignment of large biomedical ontologies. There are many issues in scaling up ontology alignment beyond a few hundred classes on each of the two ontologies being aligned. We would like to come up with various algorithms for ontology alignment and experiment into how we can work with large ontologies with thousands or tens or thousands of classes.