Each matrix block is composed as follows: (1) Consumer users??™ requirements: data
(h), query ID, quality dimensions (i), aggregations (j), global priority of the query
(from Phase II), and quality dimension priorities given by the users in Phase III;
(2) Data producer users??™ information, obtained in the previous step of this phase:
a submatrix indicating requirements fulfillment for each available data source.
According to the degree of fulfillment, a value is given (1, 3, or 9, d in Figure 2),
using the following criteria: ???1??? is given if the condition is not fulfilled, ???3??? if the
condition is not fulfilled, but can be computed from the data in the source; and ???9???
if the condition is fulfilled. For the sake of brevity we do not extend on how to de-
Data Quality-Based Requirements Elicitation for Decision Support Systems 8
Copyright ?© 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission
of Idea Group Inc. is prohibited.
termine these values. (3) Data source performance for each query (e in Figure 2);
(4) Global data source performance (f in Figure 2).
The data source performance for each query is computed as:
PerfLocal(S,Q,D).=.??‘ (prii * reli) ,
where
Figure 2. Quality assessment matrix
82 Vaisman
Copyright ?© 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of
Idea Group Inc.
Pages:
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178