Database Reference
In-Depth Information
<
:
fabg : u 11 u 12 +u 21 u 22 +u 31 u 32 +u 41 u 42 0:2
facdg : u 11 u 13 u 14 +u 21 u 23 u 24 +
+u 31 u 33 u 34 +u 41 u 43 u 44 1:2
Bd + (F 0 D )
8
<
:
feg : u 15 +u 25 +u 35 +u 45 < 0:2
ffg : u 16 +u 26 +u 36 +u 46 < 2:2
fbcg : u 12 u 13 +u 22 u 23 +u 32 u 33 +u 42 u 43 < 0:2
fbdg : u 12 u 14 +u 22 u 24 +u 32 u 34 +u 42 u 44 < 2:2
Bd (F 0 D )
<
:
u 11 +u 12 +u 13 +u 14 +u 15 +u 16 1
u 21 +u 22 +u 23 +u 24 +u 25 +u 26 1
u 31 +u 32 +u 33 +u 34 +u 35 +u 36 1
u 41 +u 42 +u 43 +u 44 +u 45 +u 46 1
T q 2D X
Fig. 16.4: The constraints in the CSP of the running example.
Table 16.4: Database D X after the solution of the CSP.
a
b
c
d
e
f
1
1
1
1
1
0
0
1
0
0
1
0
1
0
1
1
1
0
0
0
0
0
0
0
has to be ensured. Notice that if a safety margin was used, the last set of constraints
(regarding the validity of transactions in D X ) would have been excluded from the
CSP to allow for the minimization of the size of D X at a later point. After formu-
lating the CSP the CDR approach of Figure 16.2 is applied. Solving the produced
CSP using BIP creates the transactions of database D X presented in the Table 16.1.
The hiding solution of Table 16.1 is exact. Notice, that with a minimal extension
of 4 transactions, all the sensitive knowledge was hidden, leaving other patterns
unaffected in D. This solution is optimal with a distance of dist(D O ;D) = 9.
To highlight some other aspects of the hybrid hiding methodology, consider the
hiding of S =fabcg in the same database D O . In this case, using (16.1) we have
that Q = 1. However, as it turns out, only one transaction is insufficient to provide
an exact solution to the formulated CSP. Suppose that a safety margin of 3 was
used, leading to the construction of 4 transactions in database D X . Due to the use
of the safety margin in the CSP formulation, the set of constraints corresponding to
the validity of the transactions in database D X are excluded, thus, the formulation
of Figure 16.1 is used instead. Table 16.4 shows database D X after the solution
of the CSP. As one can notice, the last transaction of D X is null. Based on (16.8),
k = max(13; 0) so the empty transaction from D X can be safely removed, without
 
 
Search WWH ::




Custom Search