Discussion from Richard and Felix:

Richard Ishida: felix, where did the term 'data category' come from?  was it out there, or did we make it up?
fsasaki: Christian brought it up, saying it is defined in an ISO standard. He then had an action item to get the reference, which we forgot to keep track of.
Richard Ishida: i think DITA is using something like 'information type' - which sounds much more readily understandable
Richard Ishida: and less like a 'false friend'
Richard Ishida: data categories always has me thinking of data types
Richard Ishida: which i suppose you could justify using xml schema speak, but nontheless...
Richard Ishida: not as clear for the newcomer
fsasaki: we need a term which makes clear that we define "abstract" what e.g. translatability means (that is, we define it in prose), and have several concrete implementations: local versus global, and schema declarations in several schema languages. What term would make that difference clear?
Richard Ishida: information type
Richard Ishida: its intuitively understandable
Richard Ishida: too
fsasaki: not for me ...
fsasaki: do you have a reference except DITA? Since DITA uses "information type" quite different than what we do in ITS. also, the DITA mechanisms are quite different from our global rules
Richard Ishida: no, no other reference
Richard Ishida: it's just that data category never conveyed to me what we were talking about and needed defining and explaining and always sounds foriegn (made up), whereas 'information type' immediately suggests something to me that is a good base for my understanding
Richard Ishida: note that my reading of 'information type' is not 'information related type', but 'type of information'
fsasaki: my English native speaker feeling is too week to judge this  ; however, if you mean "type of information", it might be better to use that term. E.g.: "ITS defines various types of information: information about translatability, localization information etc. This information can be implemented in various ways (as for usage in instances: global versus local; as for ITS markup declarations: schema language specific versus independent)." . That is: we avoid any technical term for this abstract layer (no "data categories" and no "information type"), but just say "information".  
Richard Ishida: that could be an improvement