Schema Translation - Information Systems Reengineering, Integration and Normalization - page 88

5(&25'
'(3$570(17

(17,7<
'(3$570(17

(17,7<
(03/2<((

+$6

1

VHW

1

FRUUHVSRQGLQJ((5PRGHO

5(&25'
(03/2<((

1HWZRUN6FKHPD

5(&25'
6833/,(5

5(&25'
3$576

(17,7<
6833/,(5

(17,7<
3$576

6833/<
47<

1

1

VHW

VHW

1 FRUUHVSRQGLQJ((5PRGHO

1

5(&25'47<

1HWZRUN6FKHPD

Fig. 3.7 Map 1:n and m:n relationship from network to EER model

(PSOR\HH 3URMHFW 6NLOO (PSOR\HH

P Q

VHW VHW VHW

7H[W
ERRN
XVHG

Q P

6NLOOXVHG 3URMHFW Q P 6NLOO

1HWZRUNVFKHPD FRUUHVSRQGLQJ((5PRGHO

Fig. 3.8 Map n-ary relationship to EER model

Substep 6—Derive aggregation, generalization, and categorization.
An aggregation is derived if an m:n relationship from step 4 further relates to
another entity. The knowledge acquisition system should be able to detect a possible
aggregation if there is a potential m:n relationship relation record type that is further
linked to another record type. In the network schema, such a relationship can be
represented by the record type shown in Fig. 3.9 .
A disjoint generalization is derived by mapping isa relationships and their re-
cord types to a superclass/subclass entities relationship such that a superclass entity
(mapped from an owner record type) is a generalized class for the subclass enti-