2 Semantic model concepts
This chapter covers
- Modeling relationships, table expansion, and column lineage
- Exploring how report interactions translate into filter context at query time
- Navigating expanded tables and disconnected entities
- Laying the foundation for model-aware, filter-driven DAX logic
Building on the visual flow of how DAX measures are evaluated from Chapter 1, it’s clear that Power BI reports and visuals don’t query your original data sources directly. They query the semantic model, which sits between your data and the report and defines tables, relationships, and calculations. That model is not a background detail; it’s the infrastructure that governs how Power BI interprets data, applies filters, and computes results.
If you come from Excel formulas alone, this shift can take a moment to absorb. But if you have used Power Pivot, PivotTables, or Power Query, you have already worked with a model-first workflow. In Excel formulas, calculations often feel anchored to what you see on the grid.
In Power BI, DAX is part of the semantic model, and visuals display the outcome of queries evaluated over that model. Whether you are dragging fields onto a visual, building relationships, or writing measures, your work relies on the semantic model operating quietly behind the scenes.