Constraint acquisition needs better benchmarks

Current benchmarks for Constraint Acquisition (CA) and Mathematical Programming (MP) models are insufficient, impeding research reproducibility and comparability. This work introduces MPMMine, a new benchmark suite designed to validate and enhance MP models through diverse domain knowledge artifacts, promoting consistency and openness.