Normalization

Jeremy Vigil

August 24, 2002

DBM/350

Instructor

Vickie Martin

Normalization

            Normalization is removing anomalies from your data.  This can be insert, delete, or update anomalies.  Below are explanations of 3 normalization forms.  Each normalization form reduces the amount of redundancies and anomalies from the previous normalization form.

1nf

            1nf only requires you not to have any redundant columns.  This is the most basic normalization.  It is also common sense that you do not want redundant columns.  However, this only removes redundant data it still leaves the problem of having to update multiple fields for one change.  See Below Example

	Customer id	Customer name	Purchased Item	Cost
	256	John	Lawnmower	500
	257	George	Lawnmower	500
	258	John	Rake	15
	259	Hank	Rake	15
	260	Henry	Lawnmower	500
	261	Harry	Drill	50
	262	Hank	Drill	50

 

2nf

            2nf helps with redundant and update anomalies.  The below example demonstrates the update anomaly.  What if we want to sell an additional item that no one has purchased yet.  This can be chain saw.  If we add it to the above table, we would be missing customer information.  Below we can add the chain saw and price with out having to update the customer information.

	Customer id	Customer name	Purchased Item
	256	John	Lawnmower
	257	George	Lawnmower
	258	John	Rake
	259	Hank	Rake
	260	Henry	Lawnmower
	261	Harry	Drill
	262	Hank	Drill
		Purchase Item	Cost
		Lawnmower	500
		Drill	50
		Rake	15

 

3nf

3nf reduces transitive dependencies.  This is another layer added to 2nf, which further reduces the redundant data.  Below is an example from the book.

	Sid	Building	Fee
	100	Randolph	1200
	150	Ingersoll	1100
	200	Randolph	1200
	250	Pitkin	1100
	300	Randolph	1200
	Sid	Building		Building	Fee
	100	Randolph		Randolph	1200
	150	Ingersoll		Ingersoll	1100
	200	Randolph		Pitkin	1100
	250	Pitkin
	300	Randolph

 

Justification

            I would select 3nf to do normalization.  It reduces most of the redundant data and manipulation anomalies.  Although Domain key takes care of all redundant data, it would probably be too time consuming to implement.  1nf and 2nf do not reduce redundant data enough to use.

Give example