Just like in #1, the schema for ontologies used in XMP in PDF/A must be included. CiTO defines the cites relation, but FaBiO defines the classes of bibliographic objects. Possibly there are too many classes, so I'd say start with the most widely used classes.
Classes
Properties
Some properties were defined in already supported ontologies like PRISM and DC.
There should be a balance between embedding every property of a referenced item and having all the properties after dereferencing the URI of the item or finding an external description of the item.
Just like in #1, the schema for ontologies used in XMP in PDF/A must be included. CiTO defines the
citesrelation, but FaBiO defines the classes of bibliographic objects. Possibly there are too many classes, so I'd say start with the most widely used classes.Classes
Properties
Some properties were defined in already supported ontologies like PRISM and DC.
There should be a balance between embedding every property of a referenced item and having all the properties after dereferencing the URI of the item or finding an external description of the item.