A multi-driven approach to requirements analysis of data warehouse model: A case study

Abstract

In this paper, a multi-driven approach to data modeling in data warehousing will be presented, which integrates three existing approaches normally used separately: goal-driven, user-driven and data-driven; and two approaches usually not used in data warehousing field: process-driven and technology-driven. Goal-driven approach produces subjects and KPI’s (Key Performance Indicators) of main business fields. User-driven approach produces analytical requirements represented by measures and dimensions of each subject. Process-driven approach propose improvements in business processes (by using and creating subject oriented enterprise data model) to satisfy the KPI’s, measures and dimensions identified in the previous approaches. Technology-driven approach is an enabler or an obstacle to be considered in a data warehouse model. Data-driven approach is a combination of the results of previous approaches and results in a data warehouse model. By using a multi-driven approach with five stages, a layered data warehouse model more aligned with business and individual needs can be obtained. This will be illustrated by using examples of a case study.