Using Model Generation for Data Warehouse Conceptual to Physical Schema Mapping

by Nicholson, Delmer William

Abstract (Summary)
A data warehouse conceptual design is an implementation-independent representation of the contents of an organizations data repository used for analysis. The information conveyed by the design is useful to domain experts and also to database designers as a communication mechanism to validate the correctness of the design prior to implementation. In order to continue the development process, the conceptual design must be converted to a physical implementation. The conversion combines high-level abstractions from the conceptual schema with data structures required by a computer data management system. This thesis presents a semi-automated system for converting a conceptual design schema to a physical implementation. We introduce a logical model that is based on a multilevel dictionary model management approach. We extend the original research to incorporate multidimensional modeling. We demonstrate how StarER graphical schemas can be converted to the logical format. We develop a software tool that translates the logical model to a relational format that is tightly coupled to the object model of a commercially available data warehouse engine. The tool then converts the relational format to a physical implementation. The tool provides an additional relational schema that can be populated to act as the data source for the physical implementation. We describe the software system and evaluate the results of schema conversion using Microsoft SQL Server Analysis Services to demonstrate the modeled features in the resulting data warehouse implementation.
Bibliographical Information:


School:University of Cincinnati

School Location:USA - Ohio

Source Type:Master's Thesis

Keywords:data warehouse c sharp visual studio starer conceptual schema


Date of Publication:01/01/2008

© 2009 All Rights Reserved.