Managing and Querying Versions of Multiversion Data Warehouse

摘要

A data warehouse (DW) is a database that integrates external data sources (EDSs) for the purpose of advanced data analysis. The methods of designing a DW usually assume that a DW has a static schema and structures of dimensions. In practice, schema and dimensions' structures often change as the result of the evolution of EDSs, changes of the real world represented in a DW, new user requirements, new versions of software being installed, and system tuning activities. Examples of various change scenarios can be found in [1,8]. Handling schema and dimension changes is often supported by schema evolution, temporal extensions, and versioning extensions. Schema evolution approaches maintain one DW schema and the set of data that evolve in time. Temporal versioning techniques use timestamps on modified data in order to create temporal versions. In versioning extensions, a DW evolution is managed partially by means of schema versions and partially by data versions. These approaches solve the DW evolution problem partially. Firstly, they do not offer a clear separation between different DW states. Secondly, the approaches do not support modeling alternative, hypothetical DW states required for a what-if analysis. In order to eliminate the limitations of the aforementioned approaches, we propose a multiversion data warehouse.