INFO 340 Lecture Notes - Lecture 3: Associative Entity, Database Design, Data Modeling
INFO 340 Lecture 2
Lifecycle – framework by which we follow a repetitive process
- For higher probability of success
- Factors for success: on time, for the money, with projected features
o Only 29% succeed
Systems development lifecycle
- Framework for getting things done
- Different methodology for implementing framework
o Agile (software), waterfall (construction), lean (manufacturing)
- Database development process / phases
o Conceptual data modeling
▪ Discover business processes via interviews + observations
▪ Understand users and their decisions
▪ Brainstorm, then make conceptual database diagram (often full of
mistakes)
▪ Why? Determine resources and cost of the design
o Logical database design
▪ Refine conceptual diagram
▪ Ensure that each transaction can occur (be validated)
▪ Normalization
▪ Results in entity-relationship diagram (ERD) – effectively a blueprint that
any company/vendor could create
o Physical database design & definition
▪ Implement ERD
▪ Understand the platform/product to use to implement database
▪ Develop schema - create tables, logins, procedures
Entity-Relationship model (ERD)
- Entity – object that is important for business to track
o Often corresponds to row in a table
o Designated as a square/box (e.g., STUDENT)
▪ Capitalized and singular
o Boxed and circled = associative entity (e.g., one student can register for many
courses, and a course can be registered to many students)
▪ Many : Many relationships
Document Summary
Lifecycle framework by which we follow a repetitive process. Factors for success: on time, for the money, with projected features: only 29% succeed. Different methodology for implementing framework: agile (software), waterfall (construction), lean (manufacturing) Implement erd: understand the platform/product to use to implement database, develop schema - create tables, logins, procedures. Attribute property or characteristic of entity: often corresponds to field in a table, designated as a circle/orb, camelcase, singular, dotted = derived attribute (computed or calculated, double circled = multivalued attribute. Relationship a link between entities: designated as a diamond, first letter capitalized, active (e. g. , registers) Entity vs table (relation) relationship can be expressed between entities without using foreign keys. Degrees specify number of entity types involved. Cardinalities specifies how many of each entity type is allowed. Second normal form (2nf) get rid of functional dependencies. Third normal form (3nf) get rid of transitive dependencies: every property depends on the primary key.