Definitions of DataWareHousing....

Yeah... i used the term "Definitions"... there is a reason.

Mostly for every technology or domain, you can find only ONE as FATHER of that particular technology.
Coming to DataWareHousing,for this Technology... Two Fathers are there.

-->Ralph Kimball
-->W H Inmon

Why it has two technocrats? whats the reason behind it?

Yeah, The Technology is same but the WAY of APPROACH defined by one to other is deifferent.


*According to RALPH KIMBALL...
A DWH is a Relational database, which is specially designed for Business Analysis purpose, but not for Business Transaction Processing.
A DWH is designed to support Decision Making Process. (Also called as DSS(Decision Spporting System)
Only Business Managers need a DWH for Analysis of the summery of Business.

*According to W H Inmon...
A DWH is
-> Time Variant
-> Non-Volatile
-> Integrated
-> Subject Oriented

First DWH is invented by WH Inmon in 1987


Time Variant: Simple example for this one is in analysis of a company performancy, you can observe "Daily", " Weekly", "Monthly", "Quarterly"... and so on... so this is called Time variant. A DWH supports Time Variant to analyze,
where as Operational database is not much compatible.

Non-Volatile: Yeah, once data entered or a record Inserted, it is not subjected to change. Thats why called Non-Volatile. Where as in Operational Database, record will be modified.

Integrated: We can Integrated multiple databases and also multiple types of Data Sources in ETL Applications. (OACLE, SQL, DB2.....)

Subject Oriented: It can be called as Subject Oriented because all the Data elements are linked together and relevant to each other....

okay, in my next post... will go through "DATA MART"..

Keep rocking... Enjoy the Joy