Main page

Biography

Teaching

Research

Services

Other Links


Last updated: 1/02

T. Andrew Yang

Tel.: (281) 283-3835

CSCI 5333 Data Base Management System

Spring 2002 (1/14-5/6)

Week 2:  Data Modeling (Ch. 3, 4)


I. The Phases of Database Design (see Figure 3.1): 
    1. Data: requirements collection & analysis, conceptual design, logical design, physical design.
    2. Operation: functional analysis, application program design, transaction implementation.
A. The DBMS-Independent Phases:
dbmsIndependent


B. The DBMS-Dependent Phases:

dbmsDependent

  • Question: What phases are DBMS-independent?  What phases are DBMS-dependent?
  • Question: Which phase is the end of the DBMS-independent phases?  Which is the start of the DBMS-dependent phase?
  • Question: What is the relationship between the two threads (functional vs database)?
  • Question: Where does ER modeling fit in the phases?
  • Question: Where does the relational data model fit in the phases?
  • Question: At which phase does a database designer specify the database operations?
  • Question: Why does the 'physical design' phase take both the logical schema and the high-level transaction specification as the input?
  • Question: Why does the 'transaction implementation' phase take both the application program design and the internal schema as the input?

Ø Go to the Index


II. Entity-Relationship Modeling
A. Terminology
concepts
definition
example
entity


entity type


entity set


attribute


simple (atomic) attribute


composite attribute


single-valued attribute


multivalued attribute


complex attribute


stored attributes


derived attributes


key attributes





P(V)


A: E --> P(V)


value sets (domains)
V = P(V1) x P(V2) x ... x P(Vn)


relationship


relationship type


relationship set


relationship instance

Fig. 3.9
Fig. 3.10
role

Fig. 3.11
recursive relationship

Fig. 3.11
cardinality ratio


participation constraint


total (existence) participation


partial participation


structural constraint


attributes of relationship types


the weak (child, subordinate) entity types


the identifying (owner, parent, dominant) entity type


the identifying relationship


partial key


  • Question: What makes a relationship different from an attribute?
  • Question: How is a relationship between two entities represented in a relational model?  How if in an object-oriented database? HINT: See p.55.
  • Question: It is said in the book that 'weak entity types' can sometimes be represented as complex attributes.  How?  What are the trade-offs of the alternative representations?  HINT: p.60.
  • Exercise: Give a full interpretation of ALL the relationships existing in Figure 3.2.  Take into account 'cardinality ratio', 'weak relationships' and 'weak entities'.  
  • Exercise: Compare Figure 3.2 and Figure 3.15.  Which one is more precise in expressing the relationships?

Ø Go to the Index