This is a conherent paper that considers interesting extensions of the known fairness literature with two meaningful models. The authors promise to improve some aspects of the writing and from the rebuttal we expect a good outcome. A little downside is that the algorithms / techniques follow ideas from previous works.