Defining the Logical View

The logical view provides a more detailed data structure, describing the essential data attributes and their relationships in a way that remains application-neutral. It is still abstract, focusing on what data is required rather than how it will be implemented.

1. Still Usage (Application) Neutral

• The logical data model is still independent of any specific application, representing data requirements rather than the system that will manage it.

2. Identify Essential Data Attributes

• For each entity, identify its key attributes (data fields) such as unique identifiers (e.g., Order ID, Product Code), descriptive data (e.g., Customer Name, Product Description), and other relevant fields.
  • Brainstorm by data type: Consider different types of data like names, codes, dates, and measures that may be associated with the entity.
  • Analyze existing data models: Look at existing databases, forms, and reports for insights.
  • Avoid accepting the current system’s data model without question, as it may not reflect actual business needs.

3. Assign to Logical Data Types (Domains)

• Assign each attribute to a logical data type (e.g., string, date, number). Logical data types describe the kind of data each field will store without defining specific implementation (e.g., VARCHAR, INTEGER).
  • Inherit standard length for the domain if applicable.
  • Example: Names might be strings with a length of 100 characters, while dates are formatted as YYYY-MM-DD.

4. Name According to Standards

• Attribute names should follow naming conventions that are agreed upon by the organization. Names should be meaningful and consistent across the model.

5. Draft, Review, and Refine Attribute Definitions

• Define the purpose and meaning of each attribute, ensuring it’s clear to both technical and non-technical users.
  • Example: Customer Name - “The full name of the customer making the purchase.”

6. Determine Null-ability

• Decide whether an attribute can have a null value. Some fields, like a primary key, must be mandatory, while others may be optional.
  • Example: A customer’s email address may be optional, while their ID must be mandatory.

7. Identify Best (if any) Default Value

• For attributes that are optional or frequently left blank, determine the best default value (if any).
  • Example: For an optional Customer Discount Rate field, a default of “0%” could be assigned when the field is left blank.