Comparative study on data warehouse evolution techniques

Abstract

Data warehouse integrates data from various heterogeneous information sources under a unified structure to facilitate reporting & analysis done by the organizations to provide strategic information to the decision support systems. These information sources are autonomous in nature and they frequently change their data owing to transactions being carried out within the organization and may change their schema due to evolving requirements. The existing requirements are updated and some new requirements are added in order to cope up with the latest business scenarios. In fact, data warehouse never ceases to evolve. Thus, appropriate techniques should be devised in order to handle the evolving data and schema changes so that the DW can be stored in its most updated version with all types of modifications being incorporated accurately to reflect the correct form of data subject to analysis. This paper provides a comprehensive comparison of various approaches, techniques and tools being developed by various researchers in order to resolve these issues. We have examined four techniques that address the DW evolution namely schema evolution, schema versioning, temporal warehousing and view maintenance and presented a brief tabular comparison of the explored methodologies based on various parameters. © 2011 Springer-Verlag.