Notes on A Relational Model of Data...
COS 480/580: Database Management Systems
Sudarshan S. Chawathe
University of Maine
Fall 2005
These notes, cunningly disguised as questions, relate to
E. F. Codd’s classic paper: A relational model of data
for large shared data banks (Communications of the
ACM, 13(6):377–387, June 1970). Some questions are
very simple, requiring only a basic understanding of the
terms introduced by the paper. Others are more difficult,
requiring a deeper understanding of both the paper and
current database systems. Some questions require
familiarity with material that we have not yet covered in
class; these are marked with ∘ and may be skipped for
now.
- The second sentence in Section 1.2.2 claims
“[An index] tends to improve response to queries
and updates and, at the same time, slow down
response to insertions and deletions.” Is it
possible for an index to slow down the response
to a query? Justify your answer.
- ∘ Provide XML DTDs corresponding to each
of the five structures depicted on page 378. (See
Section 4.7 of the textbook.)
- ∘ For each of the five XML representations in
your answer to Question 2, provide XPath or
XQuery versions of the query described in the
first sentence on page 379.
- Briefly differentiate between the terms relation
and relationship, as used in the paper.
- Provide a relational-algebra expression for the
active domain of a database that consists of a
single relation R(A,B,C), with all attributes of
integer type.
- Provide an E-R diagram that best captures the
model outlined in Figure 3(a).
- Map the E-R model in your answer to
Question 6 to a relational schema by using the
procedure described in the textbook. Highlight
and explain any differences between the resulting
schema and the schema of Figure 3(b).
- Is it possible to define an E-R model that does
not satisfy one or both of the two conditions at
the top of the right column of page 381? Justify
your answer and reconcile it with the claim made
below those conditions in the paper.
- Compare the general name for a data item as
described in the paper with the fully qualified
names used in a current database system, such
as PostgreSQL.
- Argue for or against the following statement
from the first paragraph of Section 1.5: A
first-order predicate calculus suffices if the
collection of relations is in normal form.
- Indicate how the actions described in the fifth
paragraph of Section 1.5 may be performed in
standard SQL.
- Does UMaine’s WebDSIS system permit
symmetric exploitation (page 382)? Justify your
answer.
- Prove or disprove: The number of directed paths
needed to support symmetric exploitation of an
n-ary relation is n!.
- What are the analogs of the named and stored
sets of relations (Section 1.6)?
- Justify the claim made by Footnote 6 (page
382). Provide examples in SQL to support your
answer.
- The definition of projection in Section 2.1.2
includes the restriction n ≥ k. Comment on its
significance.
- Compare the definition of a join used in the
paper with the definition in the textbook.
- Exhibit two relations that are not joinable
although they have a common domain.
- Provide a method to test whether two relations
are joinable.
- Prove or disprove the claim made near the
bottom of the left column on page 384
(rephrased): If binary relations R, S, and T
possess points of ambiguity x, y, and z for joining
the pairs (R,S), (S,T), and (T,R), respectively,
such that xSy, yTz, and zRx, then there is a
plurality of cyclic 3-joins of R, S, and T. (We
use the notation aXb to mean (a,b) ∈ X.)
- Is the tie operator γ, defined on page 384,
expressible using the six basic relational-algebra
operators discussed in class? Justify your answer.
- Exhibit two distinct joins of the relations R and
S of Figure 12 (page 385).
- Prove the claim made in the caption of Figure
12.
- Provide two relations that have multiple
compositions. Exhibit two distinct compositions.
- What is a connection trap? Explain the term in
your own words and provide a concrete example
from a domain other than the one in the paper.
- How may the operation of restriction (Section
2.1.5) be expressed succinctly in current
relational algebra?
- What are modern analogs of the operator sets θ1
and θ2 of Section 2.2?
- Explain how current relational design theory
addresses the the redundancy in the schema of
the employee relation described in Section 2.2.1.
- Explain how current relational design theory
addresses the redundancy in the schema outlined
in the second example of Section 2.2.1. Repeat
for the weak redundancy outlined in Section
2.2.2.
- Comment on the validity of the ideas in the
last paragraph of Section 2.2.1 in the context of
current database management systems.
- Explain how a database system may induce
redundancies as suggested by the first paragraph
of Section 2.3.
- ∘ Write the simplest standard-SQL statement
that expresses the constraint described in the
second paragraph of Section 2.3. (See Chapter 7
of the textbook.)
- Comment on the expressive power of an algebraic
query language consisting of only the operators
described in this paper.