Semi-automated data-driven methods to support ontology development

An ontology is a machine-accessible representation of knowledge. It includes concepts and relationships between them, in a particular domain of interest. Ontology development is expensive, hard to automate, and requires significant efforts from both domain experts and ontologists.

In this talk, I will describe semi-automated technique to bootstrap and support the development of ontologies from biomedical literature. Starting from a small set of articles provided by the domain experts, I will show how to expand it to a corpus useful for information retrieval, and how to extract the relevant terminology and construct a semantic knowledge graph of concepts. Finally, I will discuss similarities to ontology scaffolding, and the ability to support development of other knowledge representations (e.g. semantic nets) from text-based sources.