Level: Intermediate to Advanced
In the era of big data, dimensional models serve a wide range of functions—from traditional OLAP consumption, to resource provisioning for analytics, to information distribution for self-service. This intensive course immerses you in the principles, processes, and deliverables of dimensional modeling.
You will learn must-have skills—from the design of multi-star solutions, to bridge tables, to advanced slow change processing. You will understand how these techniques apply to three worlds: data mart development, production data sets for the data science community, and tabular views for the self-service community.
Students learn to match these techniques to real-world business complexity, and to explore their impacts on BI, analytics, and data integration tools. The course exposes the reasons behind best practices, and teaches you to make pragmatic choices on design options. You will also learn how to fit dimensional modeling into varied development frameworks (including agile) and will receive templates for capturing requirements and designs.
You Will Learn
- Why dimensional design requires collaboration between modelers, business analysts, data engineers, and developers of BI products (reports, dashboards and visualizations)
- The relevance of the dimensional model in the age of big data, analytics, and self-service (including data marts, managed data sets, and tabular views)
- How to map business processes to dimensional models (stars, cubes, or data sets) and manage multiple levels of detail (process and grain)
- When to use alternatives to the basic transaction models, including periodic snapshots, accumulating snapshots, and type-specific models
- How to cope with dimensional intricacy with techniques like bridge tables, mini-dimensions, and advanced slow change responses (type 4-7 slowly changing dimensions)
- Techniques to ensure your data warehouse will scale as new subject areas are added
- How design fits into diverse development approaches including agile, and what tasks and outputs should be incorporated
This course assumes basic understanding of dimensional models (facts, dimensions, and basic slow change techniques). It is intended for anyone who contributes to data mart development, including:
- BI program and project managers
- Business analysts
- Data modelers and architects
- BI anddashboard developers
- ETL developers and data engineers