Database Reference
In-Depth Information
Table 14.1: The original database D O used in the example.
A
B
C
D
1
0
1
0
1
0
1
1
0
0
1
1
0
1
0
0
1
1
1
1
0
0
0
1
0
0
1
0
1
1
0
0
the intermediate form of database D O , which is presented in Table 14.2. Then, for
each itemset in (V [S min ), the inline algorithm generates a constraint, as follows:
1+u 52 +u 82 msup ) u 52 +u 82 0:6
(14.6)
1+u 51 msup ) u 51 0:6
(14.7)
u 51 u 52 +u 81 u 82 < msup ) u 51 u 52 +u 81 u 82 < 1:6
(14.8)
Table 14.2: The intermediate form of database D O used in the example.
A
B
C
D
1
0
1
0
1
0
1
1
0
0
1
1
0
1
0
0
u 51
u 52
1
1
0
0
0
1
0
0
1
0
u 81
u 82
0
0
Table 14.3: The three exact (and optimal) solutions for the CSP of the example.
Solution
u 51
u 52
u 81
u 82
l 1
1
0
1
1
l 2
1
1
0
1
l 3
1
1
1
0
The first two inequalities correspond to the itemsets in V , whereas the third one
regards the sensitive itemset in S (here S min = S). From these constraints, only
the third one contains products of u nm variables and therefore needs to be replaced.
Based on Figure 14.3 the inline algorithm replaces this constraint, while introducing
 
 
Search WWH ::




Custom Search