Over the last few month’s our local (South African) DAMA chapter has been running a number of sessions discussing data modelling.
Even in these days of “unstructured data” it seems that data modelling is needed more than ever.
Why do we need a data model?
Strategically, a good data model will help us to determine which data is needed to support key business processes and where that data resides.
This in turn might help us to identify gaps in our data store – where data that is identified as necessary cannot be mapped to a physical system.
Alternatively, the data model can help us to identify redundant data – data that is stored more than once in different systems and that may leads to quality problems.
Tactically, data models help us to identify and categorize data needed for a specific purpose – such as a reporting requirement. In this case, the data model is the medium by which data requirements are communicated between business and IT.
These abilities to identify and categorize important data remain very relevant in the big data world – some may argue more than ever as volumes of data increase and threatn to become even more overwhelming.
What makes a data model good?
A data model must accurately reflect user data requirements and understanding of data.
But in itself this is not enough!
Good data models will translate t a well structured physical data store that helps to ensure referential integrity and data quality.
A good data models will be flexible and open to adaptation to meet multiple purposes and reduce the number of stores need to meet differing purposes.
More simply – a good data model will result result in a good application, while a bad data model will result in an application that does not meet the business need.
The Data Model Scorecard®
Steve Hoberman’s Data Model Scorecard ® is a tool that applies Q&A to identify both strengths and areas for improvement on your conceptual, logical, and physical relational or dimensional data models.
Talk to us about taking the online course through our partner, elearningcurve.com, to learn how Steve’s model and approach can help you to assess and improve the quality of your data models – and support you on your path to becoming a Certified Information management Professional in Data Modelling
