Vassiliadis,
Bouzeghoub, and Quix (1999) also use GQM, but in this case for identifying metrics
that allow evaluating the quality of a data warehouse once it has been developed.
Closer to our proposal, Winter and Strauch (2003, 2004) introduced a demand-driven
methodology (i.e., a methodology where end users define the business goals) for
data warehousing requirement analysis. They define four steps where they identify
users and application type, assign priorities, and match information requirements
with actual information supply (i.e., data in the data sources). There are several differences
with the methodology we present here. The main one resides in that our
approach is based on data quality, which is not considered in the mentioned paper.
Moreover, although the authors mention the problem of matching required and supplied
information, they do not provide a way of quantifying the difference between
them. On the contrary, we give a method for determining the data sources that best
match the information needs for each query defined by the user. Paim and Castro
(2003) introduced DWARF, a methodology that, like DSS-METRIQ, deals with
functional and nonfunctional requirements. They adapt requirements engineering
techniques and propose a methodology for requirements definition for data warehouses.
For nonfunctional requirements, they use the extended-data warehousing
NFR Framework (Paim & Castro, 2002).
Pages:
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147