Theory of informal intentional data model is occurring, the reality is the informal, is more informal than you thought - and not many truly understand what a data model is... along with first data principles.
The AI-gen images have come a veeery long way, haven't they? :)
"Focusing only on tools and technology." - I was thinking of a different perspective than the one you wrote. Along the lines of those who do modeling being focused on the tools and tech and what they produce. They're guided by the tools rather than using the tools to guide the process of modeling. Tools-driven modeling.
You have several good examples of the reasons data modeling efforts fail to deliver hoped for results, but I feel you’ve not fully considered the nature of the core issue – what benefit is delivered to the enterprise that undertakes a comprehensive data modeling project? Is data modeling a solution looking for a problem?
One of my clients used Salesforce™ to manage its sales and marketing efforts, SAP™ to coordinate its operations, and J.D. Edwards™ to manage its financial dealings. The client was considering retiring the J.D. Edwards™ system and moving the financial data and functionality to SAP™ as a cost-saving effort.
To do this, I knew I needed to migrate data and set up the equivalent of current J.D. Edwards™ reports in SAP™, so I created a thesaurus.
I started by cataloging the reports produced by J.D. Edwards™ that were currently in use and then built backward to understand how the data in the reports was captured or created. Then I looked for corresponding reports and data available with SAP™.
Did I need to create a set of Knowledge Graphs to do the job in front of me? Should my client have invested the time and money to create detailed taxonomies encompassing Salesforce™ and J.D. Edwards™?
I’m a believer in the value of models, but it often seems like data modeling today is working on improvements to AI-controlled flyswatters.
Are you advocating, the construction of knowledge graphs and taxonomies for business data between these well known and often highly customizable systems? Or are you saying that approach is part of the rush to AI for data management and data modeling in particular? Or something I misunderstood?
wonderful write-ups; very much looking forward to checking out each update!! this one, specifically, struck a chord with me, as i love philosophy of science +/– statistics.
perhaps you could add a bit that reflects Paul Meehl's concerns with or philosophy of [psychological] science? he does *very* well breaking things down, particularly...
(a) Meehl, P. E. (1967). Theory-testing in psychology and physics: A methodological paradox. Philosophy of Science, 34(2), 103–115.
(b) Meehl, P. E. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft psychology. Journal of Consulting and Clinical Psychology, 46(4), 806–834.
(c) Meehl, P. E. (1990). Appraising and amending theories: The strategy of Lakatosian defense and two principles that warrant it. Psychological Inquiry, 1(2), 108–141.
(d) Meehl, P. E. (1990). Why Summaries of Research on Psychological Theories are Often Uninterpretable. Psychological Reports, 66(1), 195–244.
it is wild that 58 years later, we are still debating the same things. i wish people were taught in accord with your post. again, excellent work.
I think the memory of data modeling efforts and stories told about great failures and successes plays a role in who is onboard for the process. And also how practical people feel the process is and how it will help their cause it key too. Nobody cared about the difference about raw vs curated data in a large contact center because vendors teach - we’ve solved your operational or analytical reporting needs… which is rarely the case… because you need to be able to think about events and sessions, cross channel correlation, etc
It's more art than science. What about an iterative approach where you go for the highest impact solution and iteratively refine the model. More like software engineering. Start with v0 and iterate. Of course, this will be political. But faster to see results and can alleviate some of the anxiety competing teams have.
Theory of informal intentional data model is occurring, the reality is the informal, is more informal than you thought - and not many truly understand what a data model is... along with first data principles.
The AI-gen images have come a veeery long way, haven't they? :)
"Focusing only on tools and technology." - I was thinking of a different perspective than the one you wrote. Along the lines of those who do modeling being focused on the tools and tech and what they produce. They're guided by the tools rather than using the tools to guide the process of modeling. Tools-driven modeling.
In theory, there's no difference between practice and theory, but in practice, there is... - Yogi Berra
Joe –
You have several good examples of the reasons data modeling efforts fail to deliver hoped for results, but I feel you’ve not fully considered the nature of the core issue – what benefit is delivered to the enterprise that undertakes a comprehensive data modeling project? Is data modeling a solution looking for a problem?
One of my clients used Salesforce™ to manage its sales and marketing efforts, SAP™ to coordinate its operations, and J.D. Edwards™ to manage its financial dealings. The client was considering retiring the J.D. Edwards™ system and moving the financial data and functionality to SAP™ as a cost-saving effort.
To do this, I knew I needed to migrate data and set up the equivalent of current J.D. Edwards™ reports in SAP™, so I created a thesaurus.
I started by cataloging the reports produced by J.D. Edwards™ that were currently in use and then built backward to understand how the data in the reports was captured or created. Then I looked for corresponding reports and data available with SAP™.
Did I need to create a set of Knowledge Graphs to do the job in front of me? Should my client have invested the time and money to create detailed taxonomies encompassing Salesforce™ and J.D. Edwards™?
I’m a believer in the value of models, but it often seems like data modeling today is working on improvements to AI-controlled flyswatters.
John Boddie
Are you advocating, the construction of knowledge graphs and taxonomies for business data between these well known and often highly customizable systems? Or are you saying that approach is part of the rush to AI for data management and data modeling in particular? Or something I misunderstood?
hi Joe,
wonderful write-ups; very much looking forward to checking out each update!! this one, specifically, struck a chord with me, as i love philosophy of science +/– statistics.
perhaps you could add a bit that reflects Paul Meehl's concerns with or philosophy of [psychological] science? he does *very* well breaking things down, particularly...
(a) Meehl, P. E. (1967). Theory-testing in psychology and physics: A methodological paradox. Philosophy of Science, 34(2), 103–115.
(b) Meehl, P. E. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft psychology. Journal of Consulting and Clinical Psychology, 46(4), 806–834.
(c) Meehl, P. E. (1990). Appraising and amending theories: The strategy of Lakatosian defense and two principles that warrant it. Psychological Inquiry, 1(2), 108–141.
(d) Meehl, P. E. (1990). Why Summaries of Research on Psychological Theories are Often Uninterpretable. Psychological Reports, 66(1), 195–244.
it is wild that 58 years later, we are still debating the same things. i wish people were taught in accord with your post. again, excellent work.
I think the memory of data modeling efforts and stories told about great failures and successes plays a role in who is onboard for the process. And also how practical people feel the process is and how it will help their cause it key too. Nobody cared about the difference about raw vs curated data in a large contact center because vendors teach - we’ve solved your operational or analytical reporting needs… which is rarely the case… because you need to be able to think about events and sessions, cross channel correlation, etc
It's more art than science. What about an iterative approach where you go for the highest impact solution and iteratively refine the model. More like software engineering. Start with v0 and iterate. Of course, this will be political. But faster to see results and can alleviate some of the anxiety competing teams have.