English

On the Approximability of Multistage Min-Sum Set Cover

Data Structures and Algorithms 2021-07-29 v1

Abstract

We investigate the polynomial-time approximability of the multistage version of Min-Sum Set Cover (DSSC\mathrm{DSSC}), a natural and intriguing generalization of the classical List Update problem. In DSSC\mathrm{DSSC}, we maintain a sequence of permutations (π0,π1,,πT)(\pi^0, \pi^1, \ldots, \pi^T) on nn elements, based on a sequence of requests (R1,,RT)(R^1, \ldots, R^T). We aim to minimize the total cost of updating πt1\pi^{t-1} to πt\pi^{t}, quantified by the Kendall tau distance DKT(πt1,πt)\mathrm{D}_{\mathrm{KT}}(\pi^{t-1}, \pi^t), plus the total cost of covering each request RtR^t with the current permutation πt\pi^t, quantified by the position of the first element of RtR^t in πt\pi^t. Using a reduction from Set Cover, we show that DSSC\mathrm{DSSC} does not admit an O(1)O(1)-approximation, unless P=NP\mathrm{P} = \mathrm{NP}, and that any o(logn)o(\log n) (resp. o(r)o(r)) approximation to DSSC\mathrm{DSSC} implies a sublogarithmic (resp. o(r)o(r)) approximation to Set Cover (resp. where each element appears at most rr times). Our main technical contribution is to show that DSSC\mathrm{DSSC} can be approximated in polynomial-time within a factor of O(log2n)O(\log^2 n) in general instances, by randomized rounding, and within a factor of O(r2)O(r^2), if all requests have cardinality at most rr, by deterministic rounding.

Keywords

Cite

@article{arxiv.2107.13344,
  title  = {On the Approximability of Multistage Min-Sum Set Cover},
  author = {Dimitris Fotakis and Panagiotis Kostopanagiotis and Vasileios Nakos and Georgios Piliouras and Stratis Skoulakis},
  journal= {arXiv preprint arXiv:2107.13344},
  year   = {2021}
}