Toward a framework for ontology modularity
Zubeida Casmod Khan, Catharina Maria Keet · 2015
Dividing up data or information into smaller components---modules---is a well-know approach to a range of problems, such as scalability and model comprehension. The use of modules in ontologies at the knowledge layer is receiving increased attention, and a plethora of approaches, algorithms, and tools exist, which, however, yield only very limited success. This is mainly because wrong combinations of techniques are being used. To solve this issue, we examine the modules' use-cases, types, techniques, and properties from the literature. This is used to create a framework for ontology modularity, such that a user with a certain use case will know the type of modules needed, and therewith then also the appropriate technique to realise it and what properties the resultant modules will have. This framework is then evaluated with three case studies, begin the QUDT, FMA, and OpenGalen ontologies.