In classical information theory, the maximal correlation coefficient is used to establish strong limits on distributed processing. Through its relation to the ^2-contraction coefficient, it also establishes fundamental bounds on sequential processing. Two distinct quantum extensions of the maximal correlation coefficient have been introduced to recover these two scenarios, but they do not recover the entire classical framework. We introduce a family of non-commutative L^2 (p) spaces induced by operator monotone functions from which families of quantum maximal correlation coefficients and the quantum ^2-divergences can be identified. Through this framework, we lift the classical results to the quantum setting. For distributed processing, using our quantum maximal correlation coefficients, we establish strong limits on converting quantum states under local operations. For sequential processing, we clarify the relation between the data processing inequality of quantum maximal correlation coefficients, ^2-contraction coefficients, and f-divergences. Moreover, we establish the quantum maximal correlation coefficients and ^2-contraction coefficients are often computable via linear algebraic methods, which in particular implies a method for obtaining rigorous, computable upper bounds for time-homogeneous quantum Markov chains with a unique, full rank fixed point.
George et al. (Wed,) studied this question.