Data understanding is converted into appropriate conceptual and logical models. The organization could aim towards having one set of integrated data assets, with a range of emphasis from operational reporting through to slice & dice, predictive models and dashboards.
In order to enforce the growth of the data architecture in alignment with the business intelligence strategy, Data Relish offers a range of business-oriented data modelling methodologies such as Data Vault or Kimball Data Modelling. Business requirements are set out using the BEAM modelling process created by Laurence Corr.
Often, an appropriate data modelling tool is required which supports the creation, maintenance, re-use, versioning and collaboration of data model maps, which can easily be linked and subsetted, collectively describing the enterprise data model. At Data Relish, we like the Ellie data modelling tool, but there are others. We have also used plain Visio previously.
A Data Model will:
- Represent data accurately to produce reliable, robust reports that are correct
- Verify Data definitions at the conceptual, logical and physical levels
- Define the primary and foreign keys, relational tables and stored procedures
Data modeling allows you to conceptually represent the data. The model clarifies the association between data objects and rules, and sets it out in a business-friendly fashion. Data modeling can also be used to extract models from existing systems. Data modelling is conducted as a key part of data architecture, which can then be translated into technologies such as Microsoft’s Power BI, Analysis Services, or Tableau calculations.
The output of the Data Modelling process is a business-friendly conceptual, logical and physical model that accurately captures the business entities and relationships in the data.
At Data Relish, we follow established international standards using the DAMA-DMBoK (Data Management Book of Knowledge) for the Data Strategy service.