Step 1: List candidate classes and relationships

Purpose: Use textual analysis to list the candidate classes as well as their attributes and responsibilities. Determine candidate classes using noun extraction. Discover responsibilities and collaborators from the verbs.

Review both functional requirements and use case description. Underline all of the nouns and verbs. Analyze the words to list candidate classes and relationships. Usually nouns indicate the classes and verbs indicate responsibilities.

Some obvious candidate classes are CD, Customer, Order. You should be able to discover as many as 23 candidate classes if following the noun extraction method.

You should be able to find about 16 collaborations between candidate classes. Some examples are given below.

Collaborations/Relationships
has(CD, Title)
has(CD, Review)
Has(CD, Category)
Compose(CD, Artist)
has(CD, CD_Additional_Information)

Step 2: Review Candidate Classes. Update the Classes and Collaborations.

Review the candidate class and collaboration lists from the last step. Decide for each candidate whether it is complex enough to be a class or it can simply be an attribute for a class. Apply the decisions to update both class and collaboration lists.

For example,

Personal_Information, Order_Information, can be modeled into the attribute set of their associated Entity classes, Customer and Order, respectively.

How would you handle the classes, ZipCode, Location, Category, Review and Artist?