Design and implementation of an ETL approach in business intelligence project

Abstract

In the modern business, a vast amount of data has been accumulated, but mass data has not been fully utilized. Therefore, it is a vital issue for the business and IT sector to explore these data with new technology, which is useful to support better business decision-making and help enterprises to increase profit and market share. BI (Business Intelligence) technologies emerge as the times require, and ETL (Extract, Transform and Load) which realizes the technical service and decision-making support plays an important part role in BI project. In this paper, the main characteristics, advantages and disadvantages in existing ETL methods are analyzed, and some factors affecting the performance of ETL are also summarized. Furthermore, an ETL approach which combines ETL tools and SQL coding was proposed and implemented based on EL-T (Extract, Load and Transform) framework. The practice and experiment results show that the proposed approach has better efficiency and applicability than other existing ETL methods. © 2011 Springer-Verlag Berlin Heidelberg.