Conceptual design of a temporal data warehouse

Abstract

In a data warehouse, time is one of the dimensions for the analysis, but it refers to the moment in which the transaction was made, it does not explain in detail how the attributes or relationship involved in the transactions change. We are proposing a multidimensional schema that includes not only the main fact of analysis but also the temporal schemas, which did not belong to a hierarchy within the dimensions, but will register the variation of certain attributes and relationships. To build this schema, we are proposing an expansion of the algorithm that will allow to build the conceptual design of a data warehouse semi-automatically from a temporal entity-relationship model. We will unite in one model the multidimensional schema and the temporal schema to allow the registration and analysis of temporal variations and queries about the multidimensional structure.