Abstract

A data warehouse is one of the best solutions to integrate information from distributed, heterogeneous data sources. Different sized companies have different types of information management systems. The data warehouses, which can integrate these data sources together to provide summary views, are critical for the business managers to make operation management decisions. Besides the data views of the data warehouses, sometimes, the managers also need lineage trace functions to get original data sources that contribute to the data views. Currently, more than 80% of retail companies are small sized businesses or business chains that consist of small sized companies. They need a data warehouse that integrates all kinds of data sources with lineage trace functions to improve their operations management. Based on the analysis of the requirements of small businesses business chains, we use the data warehouse [1, 2, 5, 7] and lineage trace [6, 9, 10] approaches to propose an information management data warehouse solution for the operation management of the small businesses/business chains. In the same time, we develop an 'Inventory data warehouse management system with lineage trace for small business chains' demo system. In this thesis we mainly focus on the implementation and demonstrating how to develop a local information management component, how to integrate remote heterogeneous data sources to a data warehouse, how to develop a data warehouse management system with the latest Java data engine technologies and the best performance/cost rate database technologies, and also show how lineage tracing is used in the data warehouse and how to implement it as well.