Understanding Data Modeling Concepts as a Data Engineer. ๐
๐๐๐ญ๐ ๐ฆ๐จ๐๐๐ฅ๐ข๐ง๐ is the process of creating a ๐๐๐ญ๐ ๐ฆ๐จ๐๐๐ฅ; a blueprint or plan for organizing business data or information.
Imagine a data model as a way to plan and structure data so that it makes sense and is easy to understand, just like arranging items on shelves in a store.
Then think of data modeling itself as deciding how things should be placed on those shelves in the store to make it easy for anyone to find what they need.
After we choose the ERD notation style we would want to use for modeling our data, we go ahead to design the data model, which is in three phases:
๐๐๐จ๐ง๐๐๐ฉ๐ญ๐ฎ๐๐ฅ ๐๐๐ญ๐ ๐ฆ๐จ๐๐๐ฅ:
This is like drawing a rough sketch of what you want to organize. For example, before building a house, you might draw a simple layout to decide where rooms should be. Similarly, a conceptual data model helps in planning what kind of information you want to store and how itโs related.
So the conceptual model is the first version of the data model and gives an overview of what the business system contains.
What we do here is to identify data that will be used. So, we get a high-level understanding and how the entities(tables) connect. The primary keys or other attributes are not shown and we mainly use it as a communication tool.
๐๐๐จ๐ ๐ข๐๐๐ฅ ๐๐๐ญ๐ ๐ฆ๐จ๐๐๐ฅ:
Once you have a rough sketch, you decide on the specifics. Going back to our house example, this would be like deciding the exact dimensions of each room and where the doors and windows should go.
In a logical data model, you get into the details of how data should be structured and connected.
Here, we add additional information to the conceptual model, giving further details of the model design. Here, the primary keys and the attributes(columns) are specified. Sometimes the data types are stated but they arenโt always strictly specified. They can be used as placeholders.
๐๐๐ก๐ฒ๐ฌ๐ข๐๐๐ฅ ๐๐๐ญ๐ ๐ฆ๐จ๐๐๐ฅ:
Now, itโs time to build the house. The physical data model is like the construction plans, specifying the exact materials and methods for building. In the context of data, itโs about implementing the structure you planned in the logical model in a specific database or system.
Here, you identify how the system will be implemented for a specific database management system like SQL server, MySQL, or Postgresql.
The physical model gives your final blueprint for the database design.
And thatโs a simplified overview of the data modeling concept.
Below, are a few popular tools you can use for your data modeling design.
๐๐: The data modeling is an iterative process, and you will keep going back to making changes, in each phase till you get what best fits the business requirements.
