The structure of a typical data warehouse Essay

A data warehouse is a subject-oriented, integrated, time-variant, nonvolatile collection of data in support of management's decision-making process. Subject-oriented means that data is organized by entity rather than by the application that uses the data. When the data from these operational databases is loaded into a data warehouse, it is transformed into subjects. Data about products appears once in the warehouse even though it might appear many times in files and databases in the operational environment.
Being integrated means that the data is stored in one place in a data warehouse even though the data originates from everywhere in the organization. Time-variant means that data in the warehouse represents data at various points in time in the past such as when the data was last backed up. This is unlike an operational application, which has data that is accurate as of the moment. Data warehouses also retain data for long periods of time.
Being nonvolatile means that the data is read-only, it can not be altered by the user. Data warehouses contain read-only data of highly consolidated and summarized data from multiple internal and external sources. A company can use data warehouses in support of their decision-making processing.
A typical data warehouse structure is shown in the figure below.. The central Sales table is called a fact table. A fact table has rows that contain condensed and summarized data. The fact table contains a multi-part primary key. The overall structure shown in the figure is a called a star schema because of its shape, having a central hub in the middle, where all the data meets.
There are several ways to not only view but create and organize data warehouses. The first being the Top-Down view which allows the relevant data that is necessary for the data warehouse to be selected.
Next is the Data source view, this view exposes the data being stored, and managed by operational system. This information may be documented with different levels of accuracy.
Lastly, the data warehouse view includes various tables. It represents the information that is stored inside the data warehouse, including precalculated totals and counts or the date and time of the data's origin.
Another aspect to data warehouse information processing would be the support of multiple tasks such as querying, statistical querying, and basic statistical analysis using tables, charts and graphs. Today low-cost web-based accessing tools that are then integrated with web browsers are being more popular. Analytical Processing supports basic OLAP operations. It generally operates on stored data. A valuable part of analytical processing is online analytical processing over information processing is the warehouse data.
Data Mining is the discovery of hidden patterns and trends then using those patterns to create analytical models, predict future data patterns, and present the results using visualizations tools.
In conclusion, creating and managing a warehousing system is a daunting task. Many different tools are available to facilitate different aspects of the process. Development tools are used to design and edit schemas, views, scripts, rules, queries, and reports.