Chapter 17.

Database Tables and Normalization  Eliminate repeating groups, eliminate nulls by making sure  No attribute is dependent on only portion of primary key
that each repeating group attribute contains an appropriate Conversion to Third Normal Form
Normalization data value  Data anomalies created are easily eliminated by completing
 Process for evaluating and correcting table structures to three steps
minimize data redundancies Step 2: Identify the Primary Key
o Reduces data anomalies  Primary key must uniquely identify attribute value Step 1: Identify Each New Determinant
 Works through a series of stages called normal forms:  New key must be composed  For every transitive dependency, write its determinant as PK for
new table
o First normal form (1NF)
 Determinant. Any attribute whose value determines other values
o Second normal form (2NF) Step 3: Identify All Dependencies
within a row
o Third normal form (3NF)  Dependencies can be depicted with help of a diagram
 Dependency diagram: Step 2: Identify the Dependent Attributes
 2NF is better than 1NF; 3NF is better than 2NF o Depicts all dependencies found within given table  Identify attributes dependent on each determinant identified in
 For most business database design purposes, 3NF is as structure Step 1 and identify dependency
high as we need to go in normalization process o Helpful in getting bird’s-eye view of all relationships  Name table to reflect its contents and function
 Highest level of normalization is not always most desirable among table’s attributes
o Makes it less likely that will overlook an important Step 3: Remove the Dependent Attributes from Transitive
The Need for Normalization dependency Dependencies
Example: Company that manages building projects  Eliminate all dependent attributes in transitive relationship(s)
First normal form describes tabular format in which: from each of the tables that have such a transitive relationship
 Charges its clients by billing hours spent on each contract
 All key attributes are defined  Draw new dependency diagram to show all tables defined in
 Hourly billing rate is dependent on employee’s position
Steps 1–3
 Periodically, report is generated that contains information  There are no repeating groups in the table
 Check new tables as well as tables modified in Step 3 to make
displayed in Table 5.1  All attributes are dependent on primary key
sure that each table has determinant and that no table contains
inappropriate dependencies
 Structure of data set in Figure 5.1 does not handle data very All relational tables satisfy 1NF requirements
well A table is in third normal form (3NF) when both of the following are
 The table structure appears to work; report generated with Some tables contain partial dependencies true:
ease  Dependencies based on only part of the primary key  It is in 2NF
 Unfortunately, report may yield different results depending  Sometimes used for performance reasons, but should be  It contains no transitive dependencies
on what data anomaly has occurred used with caution
 Still subject to data redundancies Improving the Design
The Normalization Process  Table structures are cleaned up to eliminate troublesome initial
 Each table represents a single subject Conversion to Second Normal Form partial and transitive dependencies
 Relational database design can be improved by converting  Normalization cannot, by itself, be relied on to make good
 No data item will be unnecessarily stored in more than one
the database into second normal form (2NF) designs
table
 It is valuable because its use helps eliminate data redundancies
 All attributes in a table are dependent on the primary key
Two steps
Issues to address in order to produce a good normalized set of tables:
Conversion to First Normal Form  Evaluate PK Assignments
 Repeating group Step 1: Write Each Key Component on a Separate Line
 Evaluate Naming Conventions
 Derives its name from the fact that a group of multiple  Write each key component on separate line, then write
 Refine Attribute Atomicity
entries of same type can exist for any single key original (composite) key on last line
 Identify New Attributes
attribute occurrence  Each component will become key in new table
 Identify New Relationships
 Relational table must not contain repeating groups  Refine Primary Keys as Required for Data Granularity
 Normalizing table structure will reduce data redundancies Step 2: Assign Corresponding Dependent Attributes  Maintain Historical Accuracy
 Determine those attributes that are dependent on other  Evaluate Using Derived Attributes
Normalization is three-step procedure attributes
 At this point, most anomalies have been eliminated Surrogate Key Considerations
Step 1: Eliminate the Repeating Groups  When primary key is considered to be unsuitable, designers use
 Present data in tabular format, where each cell has single Table is in second normal form (2NF) when: surrogate keys
value and there are no repeating groups  It is in 1NF and  Data entries in Table 5.3 are inappropriate because they
 It includes no partial dependencies: duplicate existing records
o Yet there has been no violation of either entity
integrity or referential integrity