Session 1
Stuart Rhea, Ed Barnes, Hiroyuki Morita, Conny Graumans, Roberto Garcia, Andres Ferreyra, Sarah Hovinga, Graham Mullier, , Cale Boriack, Clement Jonquet
Recap of progress so far, discussions from last time. Reminder about TC 347 meeting in Berlin, plenary(s) likely will have virtual option but meeting facility is not set up well for remote attendance, should prioritize in-person participation if at all possible
Types of groups/use cases
Administrative
Proposal 1 for crop grouping being a “special kind of CropDefinition”
Hierarchy managed by CropRelationship
Recommend adding a flag to explicitly indicate group or not
Would have some business rules about using taxa and refinementCodes
How to manage different crops like sweet corn or grain/field corn
Same thing with wheat: winter, spring, red, white, durham
May want to make a sweeping statement for all wheats but also in some instances need a way to call out a specific type
Use the taxa for the broad category of wheat, but use refinement codes to get to a specific type without needing a specific hierarchy
Can make groups, just need good implementation guidelines to drive consistency
Proposal 2 CropGroupDefinition
Hierarchy manged with CropGroupDefinition
Assumption that a group cannot have more than one parent
Membership in a group/subgroup for an actual CropDefinition is manged by CropRelationship
There may be an obstacle in implementation for some cases where the crop group is not a concept they support so will require some mapping
e.g. a small grain crop in a combine in proposal 2 would need to work with two different crop definition objects
General notes
Do want to make sure there are not more than one way to do things/create groups
The refinements provide a standalone definition of the crop. The groups allow contextualization
Need some structure to group crop types, e.g. for authorization of crop protection products
Also want to break a crop down as far as possible to its basic charactics
Standard taxonomy, variety, winter/summer, color
Also want list to be manageable for a end user to choose from, for example in a harvester.
Need the list to meet the needs of that user but be able to map to a list that might contain individual crops and lists
Do not see this as a hierarchy but as groups of attributes that can be associated to a crop
If we do use CropDefinintions for groups (proposal 1) and can add properties to a group
Do members of a group inherit properties from the group
There are hierarchy’s and groupings that will need to be supported just because that is how people see things
Important to note we are not trying to reinvent ontologies
Trying to come up with some “elementary” particles of meaning to give digital detail on the concepts of corn, wheat, etx
With the goal of being good enough for the primary use cases: land use, what was planted/harvested, etc.
Ran out of time, thoughts to discuss more next time
Thinking in ontology terms, I see "groups" as classes, which can be defined using different kinds of building blocks defining characteristics of the crops included in that set (class), which can be related with taxa but also refinements like growing season. So it is possible to define those classes as the class of "wheat crops", or the class of "winter crops", or even another one that is defined as the intersection "wheat winter crops"... I can try to ellaborate on that for the next meeting.
These classes are also organised hierarchically, as "wheat crops" and "winter crops" are both superclasses of "wheat winter crops".
Session 2:
Stuart Rhea, Andres Ferreyra, Kenneth Irons, Asahiko Yamada, Miura Hiroko
Discussion of Proposal 1 & 2.
A Management Agency would be the mechanism for managing the list of CropDefinitions. Does that management extend to maintaining groupings? - Essentially, yes.