A semantic model is a metadata model that contains physical database objects that are abstracted and modified into logical dimensions. A semantic model is designed to present data for analysis according to the structure of the business.
After deployment, the semantic model is presented to users as subject areas, which are made up of tables, columns, and hierarchies. In the semantic model, these are mapped to the data sources that provide data to the workbooks, analyses, and dashboards that users create and consume.
A semantic model acts like a translation layer between your application and your underlying data structures. You can use this metrics-oriented data layer that the semantic model exposes directly with APIs, with embedded visualizations, or from other analytics tools to support your enterprise's advanced analytics applications.
A well-designed semantic model meets the business requirements of the stakeholders without them needing to understand the complexity of the underlying data structure. And a well-designed semantic model enable analysts to design workbooks, analyses, and dashboards to query data in the same intuitive way that users think about their business and ask business questions.
A semantic model enables you to structure data in a business-friendly way. It enables you to add business semantics to provide meaning to the data and the governance rules that secure data access.
