Data modeling has been in use for over four decades. While several styles have evolved, some significant data model flaws remain, and only our patented compatible data modeling has resolved them. All datasets designed using traditional data modeling methods are flawed since we can only instantiate disparate datasets. Are these siloed datasets what you intended with your logical data model?
So, you create a big, beautiful, logical data model where all your data entities are related. Yet, when you instantiate a group of datasets from this logical data model, the resultant datasets are disparate and unrelated. What happened? The answer is that you are not enforcing referential data integrity between the resultant datasets. When you populate these datasets with data values, the unconstrained datasets become disparate and siloed. Forming siloed datasets has always been the most impactful data model flaw.
Figure 1: Siloed Physical Data Models
In Figure 1, we use an example Enterprise Logical Data Model to form three physical data models using traditional data modeling methods. Each of the three physical data models is siloed because no relationships exist between them. Without connecting relationships, the datasets instantiated from the physical data models will also be siloed. So, what you design in your logical data models is not what you instantiate into your datasets. Compatible Data modeling is the only data modeling style that corrects the data silo problem.
Have you ever wondered why you can’t connect logical data models? How can you expect to integrate datasets if you can’t integrate logical data models? So, instead of combining multiple data models, you design a new one using the same design methodology? Are you kidding me?
The second data model flaw is the lack of logical data model interoperability. Traditional logical data models, as designed, are incomplete in that they are incompatible with other logical data models. They each lack standardized master structural metadata required to support direct data model interoperability. We standardize each logical data model by using a common set of reusable, standardized data entities. When you properly apply the standardized data entities, they encapsulate the entire logical data model. This encapsulation with standard data entities provides a compatible data context to each logical data model. The standardized data entities deliver direct data model interoperability to your logical data models. We use our specially designed Data Compatibility Standards data entities to make any logical data model directly interoperable.
When we instantiate datasets from directly interoperable data models, the resulting datasets are directly interoperable, provided referential data integrity is enforced between the datasets. These directly interoperable datasets spontaneously form a federated data fabric.
Please contact us if you are interested in more information on these and other data model flaws.