- From the response of the first student, I think that in order for us to continue expanding our databases and all of its features, the information gathered from components of relational model, integrity constraints and data normalization should be used together. This is because the information provided by these topics helps in understanding the constrains, rules, entities, attributes, and anomalies that is required in creating a good database. In creating a database with proper tables, rows and columns, relational model is required and it helps in correct manipulation of data, and implementing business rules that maintain integrity of the manipulated data. Similarly, integrity constraint provides entity integrity and referential integrity to ensure any foreign key value within the created database matches the primary key value in the relation of the one side. Finally, in order to validate and improve a logical design that prevents data duplication within the database, data normalization information is required.
- From the response of the second student, I think that a good database design required to store an invoice should have an invoice ID primary key containing no null values. Moreover, fully functional dependencies for such a database should include attributes that provides a relationship with another attribute. The transitive dependencies for this database should be the invoice and product. In normalizing the table to 2NF, invoice relates to products, and invoice ID relate to product name. On the other hand, in normalizing the table to 3NF, invoice relates to product, and product to customer. Invoice ID relate to product name, and product name to Customer ID. De-normalizing the tables can also be done when adding back redundant data after the tables have already been normalized in order to allow normalization to dictate which table they belong to.
