Data Ware House Concepts
Data ware housing is a Relational data base it has its own characteristics.
Integrated Data Base TINS
(1) Time Variant:
Data ware house is a time variant data base source, the business users perform analysis on their business information with respect to various time period.
From the above example the business users or business analytics can analyse their business information with respect to year wise or Quarter wise or month wise or week wise or Day wise as well.
(2) Integrated Data base:
Data ware house is built by integrate the data various operational sources into single data base.
From the above example integrating the data from various operational sources (Excel, OLEDB, Flat file source) into data ware house.
Simply data ware house cab be build by integrate the data from various operational sources.
These core tutorials will help you to learn the Dataware House. For an in-depth understanding
and practical experience, explore MSBI Training Certification.
(3) Non – Volatile:
Once the source data is inserted into the data ware housing it doesn’t reflect the changes since it is static or read only data.
|Policy ID||Policy Name||Customer Name||Address||Modified Dates|
|123||ULP||Rama||LIG – 13||2009-12-12|
|123||ULP||Rama||LIG – 14||2010-10-12|
|123||ULP||Rama||LIG – 15||2010-10-13|
(4) Subject Oriented
Data ware house is a subject oriented data base and it stores specific data about specific department in the complete organization. It is also known as data mart.
Note :- Data mart is also known as HPQS (High Performance Query Structures)
Data warehousing Architectures:
In designing data models for data ware houses or data marts, the most commonly used schema types are,
- Star schema
- Snowflake Schema
The star schema data ware housing design contains at least one fact table and surrounded by dimension tables like a star tech dimension is represented as a single table. The primary key in each dimension table is related to foreign key in the fact table.
- A simple star schema consists of one fact table and a complex star schema have more than have more than one fact table.
- All measures in the fact table are related to all the dimension tables.
Learn more about MSBI Interview Questions in this blog post.
From the above diagram the fact table contains customer ID, PID, Address ID, cost and time key attributes.
Note: The different dimensions are not related to one another.
The Snow Flake schema is an extension to star schema, where each point of the star schema explodes or divides into more points. In star schema each dimension is represented by a single dimension table, where as in a snow flake schema the dimension table is normalized into multiple look up tables, each representing a level in the dimensional hierarchy.
In the above data ware housing schema example, we have three lookup tables (Dim category, Dim product sub category and Dim address type). Product sub category lookup table is connected to product category and Add type is connected to add dimension.
For an Indepth knowledge on MSBI click on: