What is the Tableau Semantics?
Tableau Semantics provides business context and a single source of truth for data in Salesforce Data 360 by enabling a flexible and governed place for our users to create and manage metrics, dimensions, relationships, and goals.
Semantic Models are first class Salesforce metadata, integrated throughout Data 360 to power analytical and data driven experiences. Semantic Models are a grouping of Semantic Definitions tailored for a specific analytical use case. Semantic Models must consist of one or more data objects.
Semantic Models act as the layer where users can create new Semantic definitions, or override and enrich semantics that originate from the data lake layer. For example, renaming fields to provide more accessible terminology to the Business Users, overriding Relationship definitions to support a specific use case, or creating a new Metric.
Semantic Definitions are metadata that captures semantic information. Each Semantic Definition is a different type of Semantic metadata that can be utilized by different applications. Semantic Definitions are reusable and composable across applications and analytics. Semantic Definitions could be extended or overridden at different levels. The goal is to allow flexibility in using Semantic Definitions, while keeping a mutual baseline of Semantics across the organization. Tableau Semantics provides centralized experiences to create Semantic Definitions that are composable and reusable across Semantic Models and other Data 360 concepts.
The primary definition types are:
- Dimensions
- Calculated Dimensions
- Measures
- Calculated Measures
- Relationships
- Logical View
- Metrics
Dimensions are attributes of data. They are used as labels to describe measures, and allow data to be segmented or grouped. A Dimension can be Basic or Hierarchical.
A Dimension has additional metadata such as description, Data type (datetime, date, geo), and more.
Calculated Dimensions expressions defined by a set of formulas and other dimensions or measures. Calculated Dimensions can be used to classify, group, and apply other operations on Dimension values.
A Calculated Dimension has additional metadata such as description, Data type (datetime, date, geo), and more.
Measurements are a numerical attribute of data, and can be broken down by one or more dimensions.
A Measurement has additional metadata like Default aggregation, Data type (number, currency, duration), aggregation type, and more.
Calculated Measurements are expressions that are built based on measures, dimensions, and formulas. Typically, Calculated Measurements are based on dimensions and measurements originating from different source tables, and are calculated in run-time according to a specific query context.
A Calculated Measure has additional metadata like Default aggregation, Data type (number, currency, duration), aggregation type, and more.
Relationshipsare a flexible connections between different tables based on common fields, designed to keep the data's granularity intact. Relationships can be physical or logical, and have additional metadata like Join Criteria, Cardinality, and more.
Metrics are business KPIs that can be tracked over time. They are used to track, compare, and assess performance or business processes. A metric is derived from one or more measures, and in some cases will have a certain behavior that should be applied across all queries.
Join View is a data object that contains multiple tables which are connected to each other via specific join types.