2. It replaces a relation with a collection of
smaller relations.
It breaks the table into multiple tables in a
database,
It should always be lossless, because it
confirms that the information in the original
relation cab ne accurately reconstructed
based on the decomposed relations.
3. There are two types of decomposition
methods:
i) Lossless decomposition
ii) Dependency preserving
4. Let R be a relation Schema and let f be a set
of functional dependencies on R.
Let R1 and R2 from a decomposition of R.
R1 ᴧ R2R1
R1 ᴧ R2R2
The attribute common of R1 and R2 must
contain a key for either R1 and R2.
5. Dependency preserving if every functional
dependency in R can be logically derived from
functional dependencies of R1 and R2 i.e. ,
(F1UF2)+=F+
Given a relation schema R<S,F> where F is the
associated set of FDs on the attributes in S. R is
decomposed into the relation schemas R1,R2…Rn
with FDs F1,F2,…..Fn. Then this decomposition of R
is dependency preserving if the closure of F`
(where F`=F1UF2U…U Fn) is the identical to F+
(i.e. F`+
=F+)
6. Let R(A,B,C) and F={AB} then decomposition
of R into R1(A,B) and R2(A,C).
This decomposition is lossless because the
FD {A B} is contained in R1 and the
common attribute A is a key of R1.
7. Let R(A,B,C) and F= {AB}.
Decomposition of R into R1(A,B) and R2(B,C)
is not lossless because the common attribute
B does not functionally determine either A or
C i.e. it is not a key of R1 and R2.
8. R(A,B,C,D) with the FDs F= {AB, AC,CD},
consider the decomposition of R into
R1(A,B,C) with the FD F1= {AB, AC} and
R2(C,D) with the FD F2= {CD}.
In this decomposition all the original FDs can
be logically derived from F1 and F2, hence
the decomposition is dependency preserving .
Also the common attribute C form a key of
R2. The decomposition of R into R1 and R2 is
lossless.
9. Given R(A,B,C,D) with the FDs F= { AB, A
C, AD}
Decomposition of R into R1(A,B,D) and
R2(B,C) with FDs F1= {AB,AD}, F2={} is
lossy because the common attribute B is not a
candidate key of either R1 or R2.
The FD AC is not implied by any FDs in R1
or R2 Thus, the decomposition is not
dependency preserving.