Gill and Rao (1996) classify these kinds
of systems as (a) data-driven, which emphasizes access and manipulation of large
structured databases; (b) model driven, which emphasizes the access and manipulation
of a model; (c) knowledge driven, which recommends actions to the managers,
often customized for a certain domain; and (d) document driven, integrating a variety
of storage and processing technologies. A DSS is made up of: (a) database (typically
a data warehouse); (b) components for data extraction and filtering, used to extract
and validate the data taken from the operational databases; (c) query tools; and (d)
presentation tools. A data warehouse gathers data coming from different sources of
an organization (Chaudhuri & Dayal, 1997). Data warehousing involves a series of
processes that turn raw data into data suitable to be queried. A set of data transformation
processes denoted ETL (Extraction, Transformation, Loading) exports data
from the operational databases (generally in heterogeneous formats), and after some
depuration and consolidation, load them into the data warehouse. OLAP (online
analytical processing) tools are used for querying the warehouse.
System development involves three clearly defined phases: design, implementation,
and maintenance. However, in the development cycle of traditional software system,
activities are carried out sequentially, while in a DSS they follow a heuristic process
(Cippico, 1997).
Pages:
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144