ABSTRACT
This chapter outlines a combined approach to achieve this, taking advantage of two technologies, the mathematical approach of formal concept analysis and the semantic web technologies of the web ontology language. It discusses the approaches that yield a different, and data driven, approach to ontology creation. In a posteriori ontology engineering, one starts with no preconceived concepts. The chapter examines a process to achieve a posteriori ontology engineering. Data-driven, a posteriori ontology engineering does not replace the a priori ontological construction of concepts from the empirical realization of individuals to classes. Data-driven, a posteriori ontology engineering attempts to enable this by allowing the data to define the conceptualization, rather than vice versa. The truth is perhaps at neither extreme, but the results suggest that a more empirical process can yield a more robust and actionable knowledge representation. Biological ontologies are recalcitrant to the standards and best practices of the World Wide Web.
