Algebraic versus probabilstic independence in data bases
François Bancilhon, Nicolas Spyratos · 1985
The probabilistic independence of data base mappings is introduced and studied.The evolution of a data base is simulated in terms of a probabilistic model.The occurrence of any legal data base value is assumed to obey a probability distribution and thus any data base mapping is modelled as a random variable.The probabilistic independence of data base mappings is therefore defined and compared with the algebraic independence.The probabilistic independence is in general a stronger condition than the algebraic independence.The relationship of the two notions of independence is studied and necessary and sufficient conditions are given under which the algebraic and the probabilistic independencies are equivalent.