Relational database is a term used to refer to a set of entities that are interconnected in a database scheme. In specific cases, entities can be tables or files. It is very important that students understand well what an entity is. We will discuss this specifically in other articles. This time we will discuss one important stage in database design, namely normalization.
The main purpose of a relational database needs to be normalized, more precisely its normality is tested to ‘prove’ that the database that has been compiled can be said to be ‘good’. Fatansyah (2012) states that in order for a database to be made to be categorized as ‘good’ it must begin with the design process. We can use object-oriented database modeling (Utomo, 2010), data normalization (Fatansyah, 2012), or The Entity Relationship Model (E-R Model) (Silberschatz, et al. 2011) to start database design.
The three approaches basically state that at the beginning of database design it is very important for us to fully understand the business processes of the object system that we are compiling. The ability to extract key business processes and extract basic entities greatly determines our success in producing a good database.
Three things that are emphasized are finding the basic entity, determining the relationship between basic entities (cardinality value), of course, based on the business processes that have been set at the beginning, and determining the attributes of each basic entity or attributes that may arise due to the relationship of two or more entities.
Silberschatz, et al. (2011) states that when we produce an Entity Relationship Diagram (ERD) that is designed using the E-R Model approach, the base scheme can be directly transformed directly from the E-R Model. It’s just that the resulting database schema will depend on how good our E-R Model is. Fatansyah (2012) uses the term when transformation from E-R models to database schemes should be done “carefully”.
The database is an organized collection of data, the whole is stored and accessed electronically from a computer system. Where databases are more complex, they are more developed using formal design and modeling techniques.
A database management system (DBMS) is software that supports end users, applications and basic data itself to collect and analyze data. The DBMS software also includes core facilities provided to manage the database. The total number of databases, DBMS and related applications can be referred to as “database systems”. Often the term “database” is also used to reject a DBMS, database system or application associated with a database.
The existing DBMS provides various functions that enable the management of databases and their data which can be classified into four main functional groups:
- Data definition – Creation, modification and deletion of definitions that determine the organization of data.
- Updates – Insertion, modification and deletion of actual data. 
- Retrieval – Provides information in a form that can be directly used or for further processing by other applications. The retrieved data can be available in basically the same form that is stored in a database or in a new form that is obtained by changing or combining existing data from the database. 
- Administration – Register and monitor users, enforce data security, monitor performance, maintain data integrity, handle concurrency control, and recover information that has been damaged by several events such as unexpected system failures. 
Computer scientists can classify database management systems according to the database model they support. Relational databases became dominant in the 1980s. Data models as rows and columns in tables, and most use SQL to write and request data. In the 2000s, non-relational databases became popular, referred to as NoSQL because they used different query languages.
In a number of cases in the student final project that I tested, I got this impression, that the E-R model transformation process that was done mostly did not work well, maybe because it was not “careful”. This is evident from the entities contained in the E-R model that do not reflect the business processes significantly in the organization under study. For example, there are basic entities that are considered as attributes of other entities in the ERD that are generated.
As a solution, I suggest to do a database design combination between the E-R model and the use of the normalization principle as suggested by Fatansyah (2012) if indeed you do not yet have great knowledge about database design. Ransi (2019) has provided one example of the application of the ER-Model which was directly transformed in the database scheme.
- Fatansyah, 2012, Revised Edition Database, Informatics, Bandung.
- Ransi, N., (2019) “Understanding Relational Database Systems Through MySQLWokrbench”. Kompasiana.
- Silberschatz, A., Korth, H.F., Sudarshan, S., 2011, Database System Concepts, Sixth Edition, McGraw-Hill, NY, USA
- Utomo, W., H., 2010, Object Oriented Database Modeling, ANDI Publisher, Yogyakarta.