Data modeling is an exploration process to build data structures with meaningful relationships. Data modeling technically is technical representation of the business. It is therefore very important to have clear vivid understanding of the business and process in order to produce a data model that would truly serve the underlying business need. To expedite such development, at Blue Square Consulting we have developed techniques that would help collect these requirement accurately and in timely manner that may in turn help reducing the data model development cycle.
Conceptual and Logical Models
We conduct facilitated sessions with the business users for requirement gathering and validating the ideas to be able to start with the conceptual model. Once the conceptual model is complete we use the business use cases to validate the model in a facilitated session to get consensus for the user community to move forward to the next step. Once the Conceptual Model is finalized we work on detail business requirement to start building the Logical Model. At various stages of we validate the logical model with business use cases. This is logical model development in typically broken in to small iterations.
Physical Data Model
Once the logical is finalized, we move on to the next step which is building the Physical Data Model. The physical data model is built iteratively with active participation of the DBA teams. In our experience it helps to have the DBA teams involve early on in the design of the physical data model, so that the transition to the DBA team is very easy as they were part of the design process. We leverage our data architects to build out the physical data model which include the physical table structures, indexes, partitions, constraints, relationships.
