Modelo ZeEN: uma abordagem minimalista para o desenho de data warehouses

Research output: ThesisMaster's Thesis

Abstract

As the data warehouse is the core framework of a Business Intelligence system, changes to the business model at stake also imply changes to the applied data model, which require specialized maintenance and architecture operations, such as: halting the system, data warehouse redesign and reimplementation, changes to loading processes and information retrieval logic, tests, reloading of
data and system rebooting.
Considering time, risk and cost implied in these operations, strongly related to data model rigidity and complexity, it seems advisable to seek streamlining of change processes, by framing a new simple, safe and generalizable data model.
Aiming at this purpose, after reviewing existing data model concepts, and by focusing research on a specific need of the pharmaceutical industry, a new model (ZeEN - Zero Effort Entity-Network) is presented here, which was succesfully benchmarked against traditional relational and dimensional models and Anchor Modeling recent approach, for performance, and implementation and maintenance complexity.
From the experiment, conclusions are drawn over Business Intelligence generic needs, and future work is suggested.
Original languagePortuguese
QualificationMaster of Philosophy
Awarding Institution
  • NOVA Information Management School (NOVA IMS)
Supervisors/Advisors
  • Neto, Miguel de Castro, Supervisor
Award date3 Jul 2013
Publication statusPublished - 3 Jul 2013

Keywords

  • Database
  • Data warehouse
  • Data modeling
  • Business Intelligence
  • Normalization
  • Customer Relationship Management

Cite this