Normalizations remove partial dependency. All the attributesNormalizations remove partial dependency. All the attributes

Normalizations means to
be normal or convert to simple/normal form. It is usually helpful in making
data more organized. Normalization is used to remove redundancies which simply
means to eliminate unnecessary repeating things and things like partial and transitive

the database developer for a local college I would first look at the overall
school system as to how their system works. In the school system there is data
stored about students, the courses they take, their financial and personal
records, the grading records and all the other good stuff. In the upcoming
training session, I will assist the IT staff into converting table into 1NF,2NF
and 3NF. The first thing in doing so is to identify the primary key of each
table. Usually primary key is the one which is unique to each student. In some
cases, there can be even two primary keys. In this example I think the
student’s ID number or SSN could act as primary key. But another thing that can
cause problem is that some students might be international and might not have
SSN, so Student ID would be the best option which is unique to each student.
Using student ID, we can access students’ information and grades and all kinds
of related information.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

            To bring database in 1NF we must make sure there is no
repetition. No students’ information can be mixed with up with other students.
Converting to 1NF prevents these types of things as it makes the data unique
and avoids redundancy. All the attributes should be dependent on primary key and
not any other key and there should not be any type of anomalies like insertion,
update and delete. To convert data into 2NF we must remove any partial dependencies.
If there are any partial dependencies which are not dependent on primary key,
then a separate table should be made, and a primary key should be added to
remove partial dependency. All the attributes should be dependent on that
primary key. A good example of that could be if there are two primary keys in
the database and one attribute is dependent on only one primary key and not the
other then it is a partial dependency. In the students example, if both student
ID and SSN are primary key and classes which student opt for are only dependent
on ID and not SSN then it becomes a partial dependency. Another thing to note
here is that in 2NF there cannot be partial dependencies, but transitive
dependencies can exist. These types of dependencies can be eliminated by
converting data into 3NF.

            After converting data into 2NF we can convert data into
3NF by removing any transitive dependencies. A transitive dependency is the one
in which one attribute is dependent on another attribute such that it is not
part of the primary key. A transitive dependency can be removed by making a
separate table for the attribute that is creating the transitive dependency and
making it the primary key. Another important step we need to take is to make
sure that primary key of that table is set as foreign key in the original
table. A good example of this would-be student’s home address. A student home address
cannot be a primary key in the original table thus it will act as transitive
dependency, but it has things like street number, street name, apt number and
zip code dependent on it. So, to solve this problem we can make a separate
table for student address and set it as a primary key. This will convert the
data into 3NF.

            Denormalization can be defined as normalization process
in reverse. In normalization process we remove redundancies but in
denormalization we add a few extra tables or attributes. This is done in cases
where speed and performance are preferred over data redundancy. A good example
where denormalization is useful could be when in college a student wants to
change its home address. So instead of changing the name in all the different
places a new table can be created, and it can store the new updated as well as
old address. Thus, it might create data redundancy but will make the work faster
and efficient.

            Business rules affect both normalization and
denormalization in different ways. Normalization should be done to such an
extent that it gives faster results with minimal data redundancy also
denormalization should not be too much that it causes data loss. Sometimes
student details become too much handle so in that case denormalization acts as
a great source to handle data in an organized manner. Thus, business rules
should be carefully looked on and normalization should be done to an extent
that it does not the system.