At Data Modeling Zone Europe, I will be teaching a full day Advanced Data Modeling Challenges Workshop.
I always love to end these workshops by asking a few difficult data modeling questions that I have been asked over the last 12 months.
For example, at last year’s event, here are three questions that were asked:
- Explain blockchain with a simple use case (other than Bitcoin) and discuss how it will impact data modelers and business analysts.
- Can a CDM contain attributes?
- Do our current data modeling notations hamper or help with model NoSQL? Explain and provide a solution.
What questions have you faced this past year that you would like answered? (Maybe I will ask your questions at DMZ in November!)
Related Articles
1 Comment
-
2. Can a CDM contain attributes?
If by “CDM” you mean your version of a “Conceptual Data Model”, I’d say probably not.
Your CDM corresponds to the flavor of “Conceptual Data Model” that I call an “Overview Model”. In my book, “Achieving Buzzword Compliance: Data Architecture Language and Vocabulary, I describe three kinds of Conceptual Data Model:
– The Overview Model (as in your books)
– The Semantic Model (John Zachman’s Row 2), addressed by the Semantic Web and OMG’s “Semantics of Business Vocabulary and Rules (SBVR), along with E/R modeling and Object Role Modeling. This is the model that uses all the departmental language use, with all of its ugliness (hopefully resolved). Yes, in that one, attributes are included, along with described relationships.
-The Essential Model (OK, me, along with Graeme Simsioin and Graham Witt) – An abstraction of the semantic model, using patterns, to address concepts that cross departmental boundaries. Some basic attributes are included, but most are described as separate entity types: these are lists of “Parameters”.Note that all of these are “conceptual models”, in that they address the nature of the business itself, without regard to any underlying data management or other kinds of technologies.