In the context of UML-based development, it becomes
clear that the complexity of design and analysis requires impact analysis at a higher level.
In addition, model-based impact analysis allows for value to be gained before the system is
actually constructed (Briand, Labiche, O??™Sullivan, & Sowka, 2005).
Other research efforts are mapping data definitions to a single enterprise model standard. As
seen in Figure 3, we can map data definitions to a single definition or standard. For example,
suppose we define customer account as the standard for our primary account number definition.
Now each and every instance of the customer account field, irregardless of definition
or transformation rule, is mapped to this single definition. While impact analysis utilizing
this method is not as exact as the field-to-field mapping alternative, the implementation is
far simpler.
Enterpr se Metadata for Rel able Web Appl cat ons
Copyright ?© 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission
of Idea Group Inc. is prohibited.
The mapping to a single standard works equally well in the logical or model area as it does
in the physical data environment. While there are many alternatives at the data layer, moving
impact analysis across the entire asset portfolio has been problematic.
Pages:
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151