Normalizationis the procedure of correcting and evaluating a table structure toreduce data anomalies and redundancy. The process involves dividinglarge and complex tables into simpler small tables that are lessredundant. The main purpose of normalization is to isolate data andpropagate any change by the use of defined relationships (Churcher,2007). Normalized data matches to a set of rules and undergoesseveral stages known as normal forms. The first normal form, secondnormal form, and the third normal form are the most common normalforms.
Theconversion to first normal form (1NF) entails elimination ofrepetitive groups (Date, 2004). The first step involves presentingthe data into a tabular form with each cell presenting a singlevalue. In addition, there should be no repeating groups. The nextinvolves eliminating all repeating groups and nulls and ensuring thateach repeating group consists of appropriate data. The second step isto identify the primary key. This involves creation of a new key thatuniquely identifies a particular value. The third step is to identifyall dependencies with the help of a dependency diagram. Thedependency diagram is useful while depicting dependencies within atable structure and helps to identify the most important dependency.
Theconversion to the second normal form (2NF) entails elimination ofredundant data. The first step is to write down all the keycomponents on a separate line, and the original key on the last line.The second step is to determine dependent attributes and assigningcorresponding dependent attributes. The table should have no partialdependencies and dependent attribute (Date, 2004).
Theconversion to the third normal form (3NF) entails elimination ofcolumns that are not dependent on keys (Date, 2004). The first stepis to identify every new determinant. The second step is to identifythe dependent attributes on the determinants and dependency. Thenname the table with respect to its contents and functions. The thirdstep is to eradicate the dependent attributes from transitivedependencies. Then, check all the tables to ensure they havedeterminants and do not have inappropriate dependencies. Eventually,a table is in 3NF when it does not have transitive dependencies.
Ina college environment, there are several reasons for convertingdatabase tables to the first, second, and third normal forms. Forinstance, the college administration can normalize their database tomake it easy while retrieving a student’s information (Plewand Stephens, 2004).They can use the students’ admission number as the primary keysince it is unique for every student. Secondly, normalizing databasesaves space and depicts the data clearly. In addition, normalizationensures data integrity that is, every field belongs to the correctcell.
Insome typical situations, de-normalization database is acceptable.Plewet al.(2004) states that an institution can denormalize a database in casethey want to advance performance on devices such as mobile phones andcomputerised cash registers. In addition, a database may denormalizeto allow storage of redundant information in the disk by the databasemanagement system (DBMS). For instance, one can denormalize adatabase to add a field name to the table. Further, denormalizationis applicable when RDBMS does not exist and a rapid response issignificant.
Mostbusinesses chose to normalize and denormolize their databases forvarious reasons. Firstly, they do so to attain a better databaseorganisation and to reduce redundant data. Secondly, normalizingdatabase makes work easier for the database administrator and theusers of the table. According to Churcher (2007), a normalized tablehas a simplified data structure and saves on the disk space.Minimising duplicate data reduces the possibility of dataconsistency. Lastly, normalized database ensure security of the datasince some data is only accessible to certain users.
Churcher,C. (2007). Beginningdatabase design.Berkeley, CA: Apress.
Date,C. J. (2004). Anintroduction to database system(8thed.). Boston: Pearson/Addison Wesley.
Plew,R., & Stephens, R. (2003, January 24). The Database NormalizationProcess. InformIT:The Trusted Technology Source for IT Pros and Developers.Retrieved August 7, 2014, fromhttp://www.informit.com/articles/article.aspx?p=30646