Efficient maintenance of temporal data warehouses

A temporal data warehouse can be defined as a set V of materialized views over non-temporal sources. We present a self-maintainable temporal data warehouse that, besides a set of temporal views V, has two kinds of auxiliary views: a set T of auxiliary relations containing only temporal information necessary to maintain the data warehouse and a set C of complements allowing the warehouse maintenance without consulting the source databases. Several features of our approach make it especially attractive as a maintenance method for warehouses: the entire history of source databases is not stored in the warehouse; the warehouse keeps enough information for its self-maintenance; and the warehouse maintenance is achieved by executing only relational algebra operators (i.e. non-temporal operators). Although temporal views in V are specified in a temporal algebraic query language we translate them into relational algebra expressions containing only auxiliary relations in T and complements in C. We claim that our method which extends some new techniques, can simplify considerably the implementation of a temporal data warehouse.