Quick introduction to common terms for Data Warehousing and Data Modeling

The data warehouse introduces new terms and extends the Glossary for data modeling. To make the description of this article complete, I will introduce the most common terms below.

◆ Data Warehouse

A data warehouse is a collection of data that supports management decisions. Data is subject-oriented, integrated, hard-to-lose, and time variable.

A data warehouse is a collection of snapshots for all operating environments and external data sources. It does not need to be very accurate, because it must be extracted from the operating environment at a specific time.

◆ Data mart

Data Warehouses are only limited to regions with a single topic, such as customers, departments, and locations. A data mart can depend on a data warehouse when obtaining data from a data warehouse, or it does not depend on a data warehouse when obtaining data from an operating system.

◆ Fact

Fact is an information unit in a data warehouse and a unit in a multi-dimensional space. It is restricted by analysis units.

Facts are stored in a table (when a relational database is used) or a unit in a multi-dimensional database.

Each fact includes basic information about facts (income, value, satisfaction record, etc.) and is related to dimensions.

In some cases, when all the necessary information is stored in dimensions, the fact is that the data warehouse has sufficient information. We will discuss the lack of facts later.

◆ Dimension

A dimension is the axis of a coordinate system bound to a space defined by a coordinate system. The coordinate system in the data warehouse defines data units, including facts.

An example of a coordinate system is a Cartesian (Cartesian) Coordinate System with X and Y dimensions.

In a data warehouse, time is always one of the dimensions.

◆ Data Mining

The process of discovering new information in the data warehouse is called data mining, which is not obtained from the operating system.

◆ Analysis space

An Analytic space is a quantitative amount of data in a data warehouse. It is used for data mining to discover new information and support management decisions.

◆ Slices

A technology used to restrict an analysis space in a dimension to a data subset in a data warehouse.

◆ Block Cutting

A technology used to limit the analytic space of multiple dimensions to a data subset in a data warehouse.

◆ Star mode

A multi-dimensional analysis space using relational databases is called the star mode.

The star schema will be further discussed later in this White Paper.

◆ Snowflake Mode

For whatever reason, when dimensions in the star mode need to be normalized, the star mode evolves to the snowflake mode.