Parameter Sensitivity Algebra (ASP) within the Preference Propagation Framework (PPF)

This record presents a proposed formulation of Parameter Sensitivity Algebra (ASP) for extracting criterion-specific parameter elasticities from representative equations. For an additive positive model, parameter elasticity is decomposed into structural sensitivity and a context-dependent term-participation factor. The resulting elasticity matrix may subsequently be used by the Preference Propagation Framework (PPF) to propagate declared criterion orientations and weights toward a Unified Preference Representation (UPR). The formulation is presented as a theoretical and computational proposal. Exact differential identities are distinguished from local monomial representations and from empirical validation. Independent mathematical review, computational benchmarking, and domain-expert validation remain necessary. AI systems were used as analytical assistants for equation checking, counterexample exploration, terminology refinement, and documentation. Their outputs do not constitute independent scientific validation or institutional endorsement.