# Explanation of modelling choices The guiding principle for the modelling in study-mapped-to-prov.ttl is that I tried to use old classes and properties used in the PROV ontology. For example, the relationship between "data store" and "CSV data" is perhaps best thought of as the CSV data bing a relalisation of the data store (realisation being a Platform Specific Model implementation of a Platform Independent Model model or, perhaps, a realisation is a 'format' of the format-independent version...) however PROV doesn't have a realisation and the closest match I could come up with is prov:wasRevisionOf, the definition for that is "A revision is a derivation for which the resulting entity is a revised version of some original. The implication here is that the resulting entity contains substantial content from the original. Revision is a particular case of derivation.". I find this quite unsatisfactory but PROV just doesn't have any 'realizationOf' or similar propertly. DDI should perhaps subclass prov:wasDerivedFrom with a property ddi4:wasRealisationOf with perhaps the following definition: "A realization is an implementation of whereby the thing being realised contains no more information that that which was present in the original. It is likely to be used for a derivation where the derived Entity is a platform-specific expression of a platform independent Entity or perhaps a specific format of the Entity which is being derived from.". ### Author Nicholas Car