Created by: Paul Pitcher
Creation Date: Aug 16, 2025
Updated Date: Oct 14, 2025

Semantic Model

(Note: formerly known as dataset in Power BI) (Related Topic: Data Model, )
There are two kinds of semantic models
Tabular (This is what i have been using throughout my career. Data is organized in Rows and columns)
Multi-Dimensional

Semantics = When people say that is "a matter of semantics", they usually mean that the two sides are effectively saying the same thing, or that the difference between them is negligible. Semantics refers to the meanings of words and phrases in a particular context. In debates, saying that some issue is "a matter of semantics" means that those who are raising it are too stubbornly wedded to the objective theory; they only care about the words being used, and do not bother to look at the intentions of the person that they are arguing with.
My note: I think this would be like 2 people fighting over the color of a car. One person says peach one says orange.

Power BI supports different kinds of semantic models (data models).
You connect to a SSAS server and you will see all of the relationships between the tables already built for you
You can import your own data, clean it, and build relationships between the different tables
You can connect to and excel workbook model. (I haven't tried this one)

"Semantic models"(Data Models) in "Power BI service" (I think that this data model information is on the power bi service website, so that everyone in the company can use the data model that you upload with your report. Keeps everyone on the same page) as a logical layer that contains the transformations, calculations, and relationships between data sources necessary for creating reports and dashboards. These models act as a single source of truth for reports across an organization, ensuring consistency and reliability.