Technical Report MSC-2006-12

TR#:MSC-2006-12
Class:MSC
Title: Designing Data Warehouses with Object Process Methodology
Authors: Roman Feldman
Supervisors: Dov Dori
PDFMSC-2006-12.pdf
Abstract: Data warehouse modeling is a complicated task, which involves knowledge of business processes, as well as familiarity with operational information systems structure and behavior. Several modeling techniques were suggested to utilize the operational system structural or behavioral model in order to construct a data warehouse conceptual model. In this thesis, we present a feature-based technique to evaluate the existing methods of operational system based data warehouse creation, and analyze the methods according to this technique. Our analysis of these methods indicates that they are limited in their applicability to model large-scale systems, as they require acquaintance with the business processes and ability to select relevant transactional entities. In addition, these usually disregard the process perspective and require multiple unassisted manual actions, as discovering measures and relevant dimensional entities.

To overcome the limitations of existing techniques, we propose OPM-based Data Warehouse Construction (ODWC), a method based on Object-Process Methodology (OPM) for constructing a data warehouse model out of an operational system specification. OPM was the modeling method of choice primarily because it unifies all system aspects within its single view, which enables the integration of both business perspective and the system data structure. The method uses both the structural and behavioral aspects of the underlying operational system to create a multidimensional conceptual data warehouse model. Utilizing the semantic features of OPM, we present two case studies to demonstrate our method. We construct a software tool, which implements our method and allows us to apply the method on large sized systems. Consequently, we compare the ODWC method to existent methods of data warehouse creation. The proposed method was evaluated by four means: (1) a feature-based evaluation; (2) case studies that showed the feasibility of the method; (3) a real life case study; and (4) a data warehouse expert evaluation. The evaluation we performed shows that the ODWC method is the most suitable for the task of transforming operational system specification to a data warehouse model for the following reasons: (1) OPM's scaling mechanisms allow presenting the operational system and the supported business processes at varying levels of abstraction. This feature aids selection of the business process to be analyzed, and allows creating cubes at different summation levels. (2) OPM allows distinction of the business objects relevant to the business functionality. (3) OPM enables clear identification of the outcomes of a business process.

CopyrightThe above paper is copyright by the Technion, Author(s), or others. Please contact the author(s) for more information

Remark: Any link to this technical report should be to this page (http://www.cs.technion.ac.il/users/wwwb/cgi-bin/tr-info.cgi/2006/MSC/MSC-2006-12), rather than to the URL of the PDF files directly. The latter URLs may change without notice.

To the list of the MSC technical reports of 2006
To the main CS technical reports page

Computer science department, Technion
admin