Database Concepts

Normalisation §

Repeating Groups §

• A repeating group is where similar data is repeated within:
  • a single entity (e.g. multiple telephone attributes)
  • a single tuple (e.g. 2 phone numbers in one attribute)
• Data repeated across tuples is a reaping group
• 

0NF (UNF) §

• Unormalised form
  • The starting stage before getting to 1st normal form
  • UNF is all the entities and attributes before they have been normalised

1NF (First Normal Form) §

1. Each entity has a unique primary key
2. Eliminate any duplicative (repeating) attributes
   1. Create separate entities for each group of related data and identify each new entity with a primary key and copy the key from the parent table
3. (If necessary) Expand any non-atomic data
   1. Split attributes until the data cannot be broken down further
      1. Not all data is required to be broken down, unless components will need to be uniquely searched. E.g. Searching for a last name will require it to be separated out of a ‘Name’ attribute.

• A table is 1NF if:

  • every data value in an attribute is atomic
  • each record does not contain repeating data
  • each row is uniquely identifiable
    • Enforced by a primary key

• Example:

  • 
  • Where a student is enrolled in fewer than three courses, there are empty cells
    • This is wasted storage space in the database
  • There is a lot of redundant data. E.g. the course Business Studies has been typed multiple times
  • How would you find all the students on the English course?
    • You would have to search three fields
  • What would happen if a student wanted to enrol on four (or five, or even six) courses?
    • The table would have to be re-designed

•