The RAKIP Model Repository doesn't have a way to communicate which of its models can be combined (joined) together. This issue exists both on the technical level (failed executions) and the scientific level (non-sensical output).
Idea:
Add a label,tag or metadata field describing that a model or module belongs to a certain group. In the RAKIP Model Repository there could be a filter-function to only display models that belong to a certain group.
In the FSKX metadata there could be a reference(maybe doi or IRI ) to a compatible model collection.
Open questions:
collection field or array: can a model belong to multiple collections?
is it realistic to expect a modeler to fill this out when annotating a FSKX file or uploading the model to the RAKIP Model Repository?
can we add a model to a collection later? Would that not change the FSKX file?
Telemetry and linked data - possible long term project:
This could open up a use-case for our linked-data and AI efforts.
Idea: RAKIP models are registered (IRI) in a graph-database, using ontologies to describe the metadata. When a model is successfully joined with another model, FSK-Lab (or another runtime) will send (possibly an EPCIS-) event to a centralized system where it registers the successful Joining (and execution). If this happens often enough, we add a directed connection(relation) "joinable" to the two models. This can then be integrated to the RAKIP-Model-Repository with the above mentioned "filter" idea.
Problem:
The RAKIP Model Repository doesn't have a way to communicate which of its models can be combined (joined) together. This issue exists both on the technical level (failed executions) and the scientific level (non-sensical output).
Idea:
Add a label,tag or metadata field describing that a model or module belongs to a certain group. In the RAKIP Model Repository there could be a filter-function to only display models that belong to a certain group. In the FSKX metadata there could be a reference(maybe doi or IRI ) to a compatible model collection.
Open questions:
Telemetry and linked data - possible long term project:
This could open up a use-case for our linked-data and AI efforts. Idea: RAKIP models are registered (IRI) in a graph-database, using ontologies to describe the metadata. When a model is successfully joined with another model, FSK-Lab (or another runtime) will send (possibly an EPCIS-) event to a centralized system where it registers the successful Joining (and execution). If this happens often enough, we add a directed connection(relation) "joinable" to the two models. This can then be integrated to the RAKIP-Model-Repository with the above mentioned "filter" idea.