Affects Version/s: None
Fix Version/s: None
INSPIRE is an European regulation based on the ISO 19115 metadata standard with some extensions. Among other things, INSPIRE defines a controlled vocabulary to use as keywords. Note that this is just one set of vocabulary (a.k.a. thesaurus); all metadata can have many set of vocabulary. The INSPIRE vocabulary is defined here in all European languages:
Current Apache SIS implementation is a simple container class at org.apache.sis.metadata.iso.identification.DefaultKeyword. The position of keywords in metadata hierarchy can be represented as below (simplified aggregation hierarchy, not inheritance hierarchy):
The TopicCategory enumeration contains items like OCEANS, INLAND_WATERS, HEALTH, SOCIETY, etc.. A mapping from INSPIRE theme to TopicCategory is defined in Annex II and III of (which document?). Examples:
- INSPIRE "Sea regions" maps to ISO OCEANS topic category.
- INSPIRE "Energy resources" maps to ISO ECONOMY topic category.
- INSPIRE "Population distribution — demography" maps to ISO SOCIETY topic category.
We could consider implementing this mapping in Apache SIS like this: if the DataIdentification.topicCategory property is not explicitly set, then infer the topic categogries from the DataIdentification.keywords property using the above mapping. Some open questions are:
- Should we apply this rule automatically for every ISO 19115 metadata, or only for those that are identified "INSPIRE-compliant"? The INSPIRE mapping seems reasonable enough to me for considering systematic application.
- Should we systematically add those topic categories to user-supplied topic categories, or should we add them only if the user didn't specified topic categories himself?