Schema Refinement and

Normal Forms

Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke

 The Evils of Redundancy
 Redundancy is at the root of several problems associated
with relational schemas:
 redundant storage, insert/delete/update anomalies
 Integrity constraints, in particular functional dependencies,
can be used to identify schemas with such problems and
to suggest refinements.
 Main refinement technique: decomposition (replacing
ABCD with, say, AB and BCD, or ACD and ABD).
 Decomposition should be used judiciously:
 Is there reason to decompose a relation?
 What problems (if any) does the decomposition cause?

Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke

 Functional Dependencies (FDs)

 A functional dependency X Y holds over relation R

if, for every allowable instance r of R:
 t1  r, t2 r,  X(t1) =  X(t2) implies (t1) Y
= (t2)
Y
 i.e., given two tuples in r, if the X values agree, then the Y
values must also agree. (X and Y are sets of attributes.)
 An FD is a statement about all allowable relations.
 Must be identified based on semantics of application.
 Given some allowable instance r1 of R, we can check if it
violates some FD f, but we cannot tell if f holds over R!
 K is a candidate key for R means that K 
R
However, K
R does not require K to be minimal!


Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke

 Normal Forms
 Returning to the issue of schema refinement, the first
question to ask is whether any refinement is needed!
 If a relation is in a certain normal form (BCNF, 3NF
etc.), it is known that certain kinds of problems are
avoided/minimized. This can be used to help us
decide whether decomposing the relation will help.
 Role of FDs in detecting redundancy:
 Consider a relation R with 3 attributes, ABC.
• No FDs hold: There is no redundancy here.
• Given A  B: Several tuples could have the same A
value, and if so, they’ll all have the same B value!
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke
Boyce-Codd Normal Form (BCNF)

 Reln R with FDs F is in BCNF if, for all X  A in F 

 A  X (called a trivial FD), or
 X contains a key for R.
 In other words, R is in BCNF if the only non-trivial
FDs that hold over R are key constraints.
 No dependency in R that can be predicted using FDs alone.
 If we are shown two tuples that agree upon X Y A
the X value, we cannot infer the A value in
one tuple from the A value in the other.x y1 a
 If example relation is in BCNF, the 2 tuples x y2 ?
must be identical (since X is a key).
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 Third Normal Form (3NF)
 Reln R with FDs F is in 3NF if, for all X  A in F 
 A  X (called a trivial FD), or
 X contains a key for R, or
 A is part of some key for R.
 Minimality of a key is crucial in third condition above!
 If R is in BCNF, obviously in 3NF.
 If R is in 3NF, some redundancy is possible. It is a
compromise, used when BCNF not achievable (e.g., no
``good’’ decomp, or performance considerations).
 Lossless-join, dependency-preserving decomposition of R into a
collection of 3NF relations always possible.
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 What Does 3NF Achieve?
 If 3NF violated by X A, one of the following holds:
 X is a subset of some key K
• We store (X, A) pairs redundantly.
 X is not a proper subset of any key.
• There is a chain of FDs K  X A, which means that
we cannot associate an X value with a K value unless we
also associate an A value with an X value.
 But: even if reln is in 3NF, these problems could arise.
 e.g., Reserves SBDC, S C, C  S is in 3NF, but for
each reservation of sailor S, same (S, C) pair is stored.
 Thus, 3NF is indeed a compromise relative to BCNF.
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
Decomposition of a Relation Scheme
 Suppose that relation R contains attributes A1 ... An. A
decomposition of R consists of replacing R by two or
more relations such that:
 Each new relation scheme contains a subset of the attributes
of R (and no attributes that do not appear in R), and
 Every attribute of R appears as an attribute of one of the new
relations.
 Intuitively, decomposing R means we will store
instances of the relation schemes produced by the
decomposition, instead of instances of R.
 E.g., Can decompose SNLRWH into SNLRH and RW.
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 Example Decomposition

 Decompositions should be used only when needed.

 SNLRWH has FDs S  SNLRWH and R  W
 Second FD causes violation of 3NF; W values repeatedly
associated with R values. Easiest way to fix this is to create
a relation RW to store these associations, and to remove W
from the main schema:
• i.e., we decompose SNLRWH into SNLRH and RW
 The information to be stored consists of SNLRWH
tuples. If we just store the projections of these tuples
onto SNLRH and RW, are there any potential
problems that we should be aware of?
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 Problems with Decompositions

 There are three potential problems to consider:

 Some queries become more expensive.
• e.g., How much did sailor Joe earn? (salary = W*H)
 Given instances of the decomposed relations, we may not
be able to reconstruct the corresponding instance of the
original relation!
• Fortunately, not in the SNLRWH example.
 Checking some dependencies may require joining the
instances of the decomposed relations.
• Fortunately, not in the SNLRWH example.
 Tradeoff: Must consider these issues vs. redundancy.
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 Lossless Join Decompositions

 Decomposition of R into X and Y is lossless-join w.r.t.

a set of FDs F if, for every instance r that satisfies F:
  X (r)   Y (r) = r
 It is always true that r   X (r)   Y (r)
 In general, the other direction does not hold! If it does, the
decomposition is lossless-join.
 Definition extended to decomposition into 3 or more
relations in a straightforward way.
 It is essential that all decompositions used to deal with
redundancy be lossless! (Avoids Problem (2).)
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 More on Lossless Join A B
1 2
A B C 4 5
 The decomposition of R into 7 2
1 2 3
X and Y is lossless-join wrt F 4 5 6
if and only if the closure of F 7 2 8 B C
contains: 2 3
 X  Y  X, or 5 6
 X  Y  Y 2 8
A B C
 In particular, the 1 2 3
decomposition of R into 4 5 6
UV and R - V is lossless-join 7 2 8
if U  V holds over R. 1 2 8
7 2 3
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 Dependency Preserving Decomposition

 Consider CSJDPQV, C is key, JP  C and SD  P.

 BCNF decomposition: CSJDQV and SDP
 Problem: Checking JP  C requires a join!
 Dependency preserving decomposition (Intuitive):
 If R is decomposed into X, Y and Z, and we enforce the FDs
that hold on X, on Y and on Z, then all FDs that were given
to hold on R must also hold. (Avoids Problem (3).)
 Projection of set of FDs F: If R is decomposed into X, ...
projection of F onto X (denoted FX ) is the set of FDs
U  V in F+ (closure of F ) such that U, V are in X.
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 Dependency Preserving Decompositions
(Contd.)
 Decomposition of R into X and Y is dependency preserving
if (FX union FY ) + = F +
 i.e., if we consider only dependencies in the closure F + that can
be checked in X without considering Y, and in Y without
considering X, these imply all dependencies in F +.
 Important to consider F +, not F, in this definition:
 ABC, A B, B C, C A, decomposed into AB and BC.
 Is this dependency preserving? Is C 
A preserved?????
 Dependency preserving does not imply lossless join:
 ABC, A  B, decomposed into AB and BC.
 And vice-versa! (Example?)
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 1
 Decomposition into BCNF
 Consider relation R with FDs F. If X  Y violates
BCNF, decompose R into R - Y and XY.
 Repeated application of this idea will give us a collection of
relations that are in BCNF; lossless join decomposition, and
guaranteed to terminate.
 e.g., CSJDPQV, key C, JP  C, SD  P, J  S
 To deal with SD  P, decompose into SDP, CSJDQV.
 To deal with J  S, decompose CSJDQV into JS and CJDQV
 In general, several dependencies may cause violation
of BCNF. The order in which we ``deal with’’ them
could lead to very different sets of relations!
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 2
 BCNF and Dependency Preservation

 In general, there may not be a dependency preserving

decomposition into BCNF.
 e.g., CSZ, CS  Z, Z C
 Can’t decompose while preserving 1st FD; not in BCNF.
 Similarly, decomposition of CSJDQV into SDP, JS and
CJDQV is not dependency preserving (w.r.t. the FDs JP
C, SD P and J S). 
 However, it is a lossless join decomposition.
 In this case, adding JPC to the collection of relations gives us
a dependency preserving decomposition.
• JPC tuples stored only for checking FD! (Redundancy!)
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 2
 Decomposition into 3NF
 Obviously, the algorithm for lossless join decomp into
BCNF can be used to obtain a lossless join decomp
into 3NF (typically, can stop earlier).
 To ensure dependency preservation, one idea:
 If X  Y is not preserved, add relation XY.
 Problem is that XY may violate 3NF! e.g., consider the
addition of CJP to `preserve’ JP  C. What if we also
have J  C ?
 Refinement: Instead of the given set of FDs F, use a
minimal cover for F.

Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 2

 Minimal Cover for a Set of FDs
 Minimal cover G for a set of FDs F:
 Closure of F = closure of G.
 Right hand side of each FD in G is a single attribute.
 If we modify G by deleting an FD or by deleting attributes
from an FD in G, the closure changes.
 Intuitively, every FD in G is needed, and ``as small as
possible’’ in order to get the same closure as F.
 e.g., A
 B, ABCD E, EF  GH, ACDF  EG
has the following minimal cover:
A
 B, ACD  E, EF G and EF H


 M.C. Lossless-Join, Dep. Pres. Decomp!!! (in book)

Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 2
 Refining an ER Diagram
 1st diagram translated: Before:
since
Workers(S,N,L,D,S) name dname
Departments(D,M,B) ssn lot did budget
 Lots associated with workers.
Employees Works_In Departments
 Suppose all workers in a
dept are assigned the same
lot: D  L
 Redundancy; fixed by: After:
budget
Workers2(S,N,D,S) name
since
dname
Dept_Lots(D,L) ssn did lot
 Can fine-tune this:
Workers2(S,N,D,S) Employees Works_In Departments
Departments(D,M,B,L)
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 2
 Summary of Schema Refinement
 If a relation is in BCNF, it is free of redundancies that
can be detected using FDs. Thus, trying to ensure
that all relations are in BCNF is a good heuristic.
 If a relation is not in BCNF, we can try to decompose
it into a collection of BCNF relations.
 Must consider whether all FDs are preserved. If a lossless-
join, dependency preserving decomposition into BCNF is
not possible (or unsuitable, given typical queries), should
consider decomposition into 3NF.
 Decompositions should be carried out and/or re-examined
while keeping performance requirements in mind.
Database Management Systems, 3ed, R. Ramakrishnan and J. Gehrke 2