Supracompact inference operations (Q1313087): Difference between revisions

Supracompact inference operations satisfy the condition that whenever a conclusion \(x\) is drawn from a set of premisses \(A\), there exists a finite subset \(A_ 0\) of \(A\) such that \(x\) can be drawn from any subset \(X\) of \(A\) containing \(A_ 0\). Supracompactness is thus stronger than both compactness and deductivity, at least in the distributive case. It is shown that any cumulative operation \(C\) contains a cumulative and supracompact operation \(K(C)\) with which it agrees on finite sets. Moreover, \(K(C)\) inherits most of the properties that \(C\) may enjoy, like monotonicity, distributivity or disjunctive rationality. Distributive supracompact operations can be characterized by a simple functional equation as well as by semantic properties. Special attention is paid to the rational case, and it is shown that rational supracompact operations can be represented by injective ranked models.