A PROBABILISTIC RELATIONAL ALGEBRA FOR THE INTEGRATION OF INFORMATION-RETRIEVAL AND DATABASE-SYSTEMS

We present a probabilistic relational algebra (PRA) which is a general ization of standard relational algebra. In PRA, tuples are assigned pr obabilistic weights giving the probability that a tuple belongs to a r elation. Based on intensional semantics, the tuple weights of the resu lt of a PRA expression always conform to the underlying probabilistic model. We also show for which expressions extensional semantics yields the same results. Furthermore, we discuss complexity issues and indic ate possibilities for optimization. With regard to databases, the appr oach allows for representing imprecise attribute values, whereas for i nformation retrieval, probabilistic document indexing and probabilisti c search term weighting can be modeled. We introduce the concept of va gue predicates which yield probabilistic weights instead of Boolean va lues, thus allowing for queries with vague selection conditions. With these features, PRA implements uncertainty and vagueness in combinatio n with the relational model.