Database Design

Database Design

Scenario

Currently a school uses a non-computerised database system to store the following information about its students: roll class, surname, first name, guardian's name, address, student's date of birth, gender, sport house, emergency-contact telephone number, mailing name, list of subjects offered, and subjects in which the student is enrolled. The system takes up a large amount of space, has many duplicated entries some of these with mistakes and it is time consuming to search. The school intends to introduce a relational-database management system to overcome these problems.

Current school data storage and retrieval system (the pre-computer database system)

„h School staff, teachers and office staff are the users of the database.

„h Database used to track all details of students in the school.

„h Large amounts of data are stored. An individual record is required for each student.

„h Lots of duplication exists as details of subjects offered by the school are duplicated on individual student records.

„h Data is organised in alphabetical order on student names within years.

„h Data is stored in a number of filing cabinets.

Advantages

„h quick to find details of a student if year and surname are known

„h data is stored in a central secure location

„h original source material is stored in the student file.

Disadvantages

„h slow to find details of students when surname and year are not known

„h data storage takes up a lot of room

„h minimal number of staff can access records at one time

„h data is duplicated.

Stages of development of the computerised relational database system

The new computerised relational database system will be developed in the following stages:

„h analysis

„h design

„h implementation

„h maintenance

Analysis

In this stage the current system is analysed to determine the functions it performs and those that it is unable to perform. Direct and indirect users of the new system are interviewed to determine what they require from the new system. The inputs to the new system, the processes it is to perform and the outputs from the system are clearly defined. A recommendation is made about the type of new system that will be required to meet the identified needs.

During the analysis stage, the following is determined:

„h current storage space of student records needs to decrease

„h an easily update-able backup of student records is required,

„h one staff member is to be responsible for the administration of the system,

„h unnecessary duplication of data is to be avoided,

„h all teaching staff, office staff and the Principal need to have access to the database simultaneously - read access only is required

„h reports such as roll class lists, subject class lists, lists of sporting houses, student contact details and family mailing lists need to be easily generated.

To meet these requirements, a relational database system has been recommended. A relational database is "a database in which data is organised in a series of relationships, or two-dimensional tables, where the columns (attributes) represent data fields, and the rows (tuples) represent records. Linking of data between records in different files is done by means of a key field."

The relational model was chosen in preference to a flat file system as the requirements of users would not be met with a flat file. A flat file is "a type of database where all of the data is contained in one file". For all of the information to be held in one table, every student record would require data to be unnecessarily duplicated, introducing redundant data. For example, every student record would require information about the subjects the student is enrolled in and these subject details would be duplicated throughout the flat file. This duplicated data increases the space required by the computer system to process and store the data. It also leads to possible