Information Technology Reference
In-Depth Information
A Trace Metamodel Proposal Based on the
Model Driven Architecture Framework for the
Traceability of User Requirements in Data
Warehouses
Alejandro Mate and Juan Trujillo
Lucentia Research Group
Department of Software and Computing Systems
University of Alicante
{amate,jtrujillo}@dlsi.ua.es
Abstract.
The complexity of the Data Warehouse (DW) development
process requires to follow a methodological approach in order to be suc-
cessful. A widely accepted approach for this development is the hybrid
one, in which requirements and data sources must be accommodated to
a new DW model. The main problem is that we lose the relationships be-
tween requirements, elements in the conceptual models and data sources
in the process, since no traceability is explicitly specified. Therefore,
this hurts requirements validation capability and increases the complex-
ity of Extraction, Transformation and Load processes. In this paper, we
propose the first trace metamodel for DWs and focus on the relation-
ships between requirements and conceptual models. We propose a set of
Query/View/Transformation rules to include traceability in DWs in an
automatic way, allowing us to trace every requirement to the conceptual
model and further increasing user satisfaction.
Keywords:
Data Warehouses, traceability, user requirements, MDA.
1
Introduction
Data Warehouses (DW) integrate several heterogeneous data sources in multi-
dimensional structures (i.e. facts and dimensions) in support of the decision-
making process [10, 12]. Therefore, the development of the DW is a complex
process which must be carefully planned in order to meet user needs. In order
to develop the DW, three different approaches, similar to the existing ones in
Software Engineering (bottom-up, top-down, and hybrid), were proposed [21,4].
The first approach follows a bottom-up process and makes use of the infor-
mation in the data sources while ignoring the user requirements. As the schema
is not adapted to the user needs [21] the DW fails to meet the user expecta-
tions. The second approach follows a top-down process and focuses on the user
requirements while ignoring the data sources. Therefore, it is possible that some
of the user needs cannot be satisfied because the necessary data has not been
