To efficiently answer queries, datalog systems often materialise all
consequences of a datalog program, so the materialisation must be updated
whenever the input facts change. Several solutions to the materialisation
update problem have been proposed. The Delete/Rederive (DRed) and the