Representing SNOMED CT concept evolutions using process profiles

SNOMED CT is a very large biomedical terminology supported by a concept-based ontology. In recent years it has been distributed under the new release format 'RF2'. RF2 provides a more consistent and coherent mechanism for keeping track of changes over versions, even to the extent that - in theory at least-any release will contain enough information to allow reconstruction of all previous versions. In this paper, using the January 2016 release of SNOMED CT, we explore various ways to transform change-assertions in RF2 into a more uniform representation with the goal of assessing how faithful these changes are with respect to biomedical reality. Key elements in our approach are (1) recent proposals for the Information Artifact Ontology that provide a realism-based perspective on what it means for a representation to be about something, and (2) the expectation that the theory of what we call 'process profiles' can be applied not merely to quantitative information artifacts but also to other sorts of symbolic representations of processes.