Information processing protocols are typically built out of simpler parts, called primitives, and two of the most important such primitives are privacy amplification (PA) and data compression. The former extracts the truly secret part of some classical data, while the latter squeezes it into the smallest possible form. We show these tasks are dual in the setting of quantum information processing. Specifically, the tasks of PA of classical information against quantum adversaries and classical data compression with quantum side information are dual in the sense that the ability to perform one implies the ability to perform the other. The duality arises because the two protocols are connected by complementarity and the uncertainty principle in the quantum setting. Applications include a new uncertainty principle formulated in terms of smooth min- and max-entropies, which are useful in the study of one-shot protocols, as well as new conditions for approximate quantum error correction.
Equivalence of three classical algorithms with quantum side information: Privacy amplification, error correction, and data compression
