Meeting minutes
Introductions and Data Workshop
<pchampin> https://
https://
Daniel: is anyone from CatenaX (automotive subset of GaiaX) is attending the workshop?
… if you're not familiar with that activity, we can discuss separately
Pierre-Antione: we reached out to GaiaX but would certainly welcome further discussion
Pull Requests
Daniel: Carine, thank you for the namespace PR, will accept that soon
https://
Daniel: the next one is based on a vulnerability discovered by bots https://
… ran into complications, wonder if we can reopen the PR?
Carine: it should work directly
Daniel: maybe if I recommit it will check it again, otherwise will create a new PR
… earlier I wasn't in the group and think that was the cause [addressed]
Carine: it's ok, I'll merge
https://
Daniel: as there is more content to this one, will give people until the next meeting to provide feedback
… in VSS typically dot-notation to a leaf is used. I added annotation property
<caribou> done
Felix: how to handle the seed properties?
Daniel: for now I took instances out for instance row1, 2...
… rest of this PR is minor changes but do please take a look
https://
Daniel: in VSS3 we have comments so added to ontology as rdfs:comment. main point is to have dynamic instances in the tree itself. we decided at the workshop to use instances instead of classes
… this is leading us to punning. we had a restriction of the property
Felix: we could work to make the object property work again
… if we agree we can accept the PR as it would be good to have what's in VSS in VSSo
DanielA: basically having instances is beneficial when you want to point to a concept from another ontology. that was the case with dynamic vehicle properties
… I haven't worked with vehicle component concept and whether it would be better as a subclass
Felix: vehicle component is similar to VSS branch, giving some details about where the property is coming from
… so far we haven't added properties nor subclasses to existing
Daniel: right now as we're generating the information and don't see anything on top we have two choices: leave it out or add as instances
Felix: what we can do later on is introduce a subclass structure later
DanielA: what I have seen in other models is adoption is difficult if knowledge is in too many domains. keeping the model simple will help with adoption
… I would support components as instances for now
Issues
https://
DanielA: by what concept is the unit to be defined?
Daniel: anyone want to propose an idea or does someone want to work on for next time
Felix: haven't formed my thoughts yet
DanielA: I have some notes and can bring forward an idea for next time
Daniel: use slack to discuss in meantime or this github issue and tag Felix
https://
Daniel: currently we use branch structure to create unique names. the main goal is to avoid overlaps/conflicts
… a better way would be appreciated
https://
DanielA: this is making the description more consistent, using position in the vehicle. I suggested using the union as the domain
Pierre-Antione: it makes the ontology a bit more complex but the right way semantically would be to use the union approach
… another way would be to create a superclass of vehicle property or component
… this trick has the advantage of only needing RDF schema (not OWL)
… there are trade offs
DanielA: positionedAt property can be used to tie to vehicle component
Felix: I'm a bit confused
Daniel: its about position eg front left
Pierre-Antione: union is correct but a string interpretation. once you set the axiom it cannot be anything else
… if you want a more relaxed association with position then a union would be weaker and not used
Felix: understand. I would express it the same way, you want to know where you can use this property
DanielA: can you remind me the range for positionedAt property?
Daniel: the sentiment I heard was more a leaning toward union representation but can leave this issue open for further discussion
DanielA: looking at resulting ontology and query difficulty would help
Daniel: from the association perspective it doesn't change
… if someone can prepare PR with examples that would help and be used to improve the documentation
https://
DanielA: this is more about making use explicit
Daniel: that's fine, any other opinion?
Felix: I agree
https://
DanielA: this will depend on component relationship. we are using hierarchy from VSS
… root is vehicle and partOf could handle this
Felix: in the diagram it is a transitive property
DanielA: but I think the intention is to have this functionality
Felix: for me it is redundant
Daniel: when discussing relationship between vehicle and components it might indeed be redundant
Pierre-Antione: partOfVehicle is for when there is an instance of a vehicle. we want a hierarchy that does not depend on any specific vehicle
Daniel: it needs an example and more thinking
Felix: exactly