Chapter 5%3A Conceptual Modeling - KGConf/BookClub-OntologyEngineering GitHub Wiki

(Ann Clark) Section 5.3 p. 66 mentions ontology for data governance. Can you provide an example of how an ontology can be used to support data governance?

Also Section 5.3 p.71 For maintenance of external ontologies - can you explain why this is important? I assumed I would import a version of an ontology into an ontology editor and select portions to use. If I've customized or only reused certain classes, how important is the external maintenance aspect?

Another Section 5.3 - If I reuse only properties, but not the specific relationships in an existing ontology, say for a machine learning use case, I assume I'd add it as reference metadata, but may not care about the maintenance or updating of the external ontology. Does this assumption hold water?

Ch. 5 overall: Is a directed graph the optimal way to visualize an ontology regardless of use case? Can you think of any use cases that would benefit from a different type of representation or visualization?

(Marcus Noelke) Section 5.3: If FIBO is a basis but not all classes/properties are used in the project, what is the recommended practise to deal with the unused items? Leave it, delete it, in-activate it(if possible)? How to deal then with possible updates of the reused ontology?

(Marcus Noelke) Section 5.3: If two upper ontologies are used (e.g. FIBO and ISO15926-14), what is the best way to deal with overlapping concepts? In this example the concepts "Aspect" and "Location" exist in both ontologies. Would you simply make the classes equivalent-to?

(Mel McCrea) is it possible to use any kind of data in model? foe example transactional data?

(Mel McCrea) (...Not in the last chapter) but, does the reasoner affect in any way how we can model the ontology?

(Louie Rumanes, 5.7, 5.8) Are there rules of thumb for deciding between adding a property to a class description (Wine.color) and creating separate classes (Red Wine, White Wine, Rose Wine)?

(Louie Rumanes) How can causal relations be used in ontologies and can reasoners be used for causal chaining?

(Louie Rumanes) How and why are rule languages like Datalog, Spin and to some extent SHACL used in ontology design? Are they extensions for reasoner capabilities or do they ust make complex relationships more explicit and easier to understand?