Thesis (M.Sc.) - Cairo University - Faculty of Computers and Information - Department of Information System

The efficiency of data warehouse staging area could almost always be better. However, in any domain like the sales domain that we have chosen for evaluation this can be a difficult question to answer. Therefore, we directed our investigation towards two main areas. In the beginning, we improved the quality of sales data throughout the extraction and transformation stages of the raw data acquired from the operational sources and evaluate the differences after applying predefined measures and sales ontologies. Afterwards, we present an evaluation of different four techniques and their impact on data transfer to its final destination. Within this thesis we start by giving an introduction regarding data warehouse. Afterwards, we explain what Staging Area is and how it is extremely important for a well-built Data Warehouse. In the attempt to enhance the Staging Area, lots of researches have been conducted. We present a literature of different ETL ontologies researches and their impact on the Staging Area. Afterwards, we refer to the chosen sales domain for applying the enhancement approach. We then introduce our approach to enhance the ETL phases. First, we applied predefined industrial ontologies besides running a semi-automated program on the extract phase then on transformation phase. After finishing both phases we noticed that the data quality of the tables have improved like with address table

Issued also as CD