INFO20003 Chap.2 Entity Relationship Modelling
Entity Relationship Modelling
Entity relationship modelling turns a narrative into explicit claims about object identity and association. You learn to distinguish entity types from descriptive attributes, select identifiers, place facts on the relationship whose pairing determines them, and read minimum and maximum cardinalities one endpoint at a time.
The chapter also maps many-to-many relationships into associative relations and shows how assumptions should resolve genuine ambiguity without inventing a different organisation. Begin by separating entity types with continuing identity from values that merely describe them.
A candidate identifier must be unique under the organisation's rules and stable enough to support references; a value that happens to be unique in a sample is not automatically an identifier. Weak or dependent objects need the identity of an owner plus a discriminator when they cannot be identified independently.
Relationship degree records how many entity types participate, while endpoint constraints record the legal number of relationship instances for one entity instance. Read each endpoint by fixing one instance at the opposite end, and keep the minimum separate from the maximum. A minimum of zero means participation is optional; a minimum of one states an existence obligation.
Attributes belong to the object or relationship whose identity determines their value. A grade belongs to a particular enrolment, and an award date belongs to a particular technician-certification holding, because moving either value onto one participant would repeat or overwrite valid facts. Recursive relationships require role names so the same entity type can participate in different ways.
Temporal requirements also affect identity: if the same pair may be linked repeatedly over time, the pair alone cannot identify all historical relationship instances. Mapping is a useful validation exercise rather than a separate drawing ritual. A regular entity becomes a relation, a one-to-many relationship normally contributes a foreign key to the many side, and a many-to-many relationship becomes an associative relation.
The mapping must retain pair uniqueness, participation meaning and relationship attributes. Some minimum-participation rules cannot be guaranteed by a simple foreign key, so a complete answer identifies the limitation and the additional enforcement required. Finally, test a model using tiny legal and illegal populations.
If it rejects a permitted state or admits a forbidden one, revise the structure or the stated assumption before polishing the notation.
What this chapter covers
- 01
Entity types
- 02
Identifiers
- 03
Attribute placement
- 04
Relationship meaning
- 05
Minimum participation
- 06
Maximum cardinality
- 07
Associative entities
- 08
Assumption discipline
Worked example · free
Model professional certifications
- 1Create Technician and Certification as independently identified entity types.
- 1Connect them through a many-to-many Holds relationship.
- 1Place award and expiry dates on Holds because they vary by pair.
- 1Map the relationship to an associative relation with both foreign keys.
- 1State any unresolved minimum participation as an assumption.
Key terms
- Entity type
- A class of independently identifiable objects about which the database stores facts.
- Identifier
- An attribute or minimal attribute set that distinguishes entity instances.
- Relationship
- A meaningful association among instances of entity types.
- Participation
- The minimum number of relationship instances in which an entity instance may occur.
- Cardinality
- The maximum number of relationship instances permitted at an endpoint.
- Associative relation
- A relation created to represent a many-to-many relationship and its attributes.
Entity Relationship Modelling FAQ
When should time become part of relationship identity?
Include a temporal discriminator when the same participant pair may form more than one valid relationship instance over time and history must remain distinguishable. If only the current relationship is permitted, the pair can remain unique. State that business rule before choosing a composite key or introducing a separate relationship identifier.
How do I stop reversing cardinalities?
Hold one instance of the entity on the opposite side fixed, then count how many partners it may have at the endpoint being labelled. Translate the completed symbols back into two ordinary business sentences.
When does a fact belong on a relationship?
Place it on the relationship when its value depends on a particular pairing. A booking timestamp depends on both the member and workshop, so storing it on either entity alone would lose the distinction among bookings.
What makes an assumption useful?
A useful assumption resolves something unstated that changes the structure, such as whether participation is optional. It should not add departments, roles or processes that the narrative never requires merely to make the model look richer.
Exam move
Model short narratives from unfamiliar domains and annotate every structural choice. Underline candidate entities, circle possible identifiers and convert verbs into relationships. For each endpoint, write a separate sentence containing the minimum and maximum. Ask which identifier set determines every relationship attribute.
After drawing, map the diagram into relations; an awkward mapping often reveals an unresolved identity or cardinality mistake. Compare your answer with a checklist instead of a memorised picture: complete identities, meaningful verbs, correctly directed cardinalities, relationship attributes at the full determinant, and assumptions limited to genuine gaps.
Working through Entity Relationship Modelling in INFO20003? Sia is AskSia’s AI Computer Science tutor — ask any INFO20003 Entity Relationship Modelling question and get a clear, step-by-step explanation grounded in how INFO20003 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.