English

APX-Hardness and Approximation for the k-Burning Number Problem

Computational Complexity 2021-01-19 v2 Data Structures and Algorithms

Abstract

Consider an information diffusion process on a graph GG that starts with k>0k>0 burnt vertices, and at each subsequent step, burns the neighbors of the currently burnt vertices, as well as kk other unburnt vertices. The \emph{kk-burning number} of GG is the minimum number of steps bk(G)b_k(G) such that all the vertices can be burned within bk(G)b_k(G) steps. Note that the last step may have smaller than kk unburnt vertices available, where all of them are burned. The 11-burning number coincides with the well-known burning number problem, which was proposed to model the spread of social contagion. The generalization to kk-burning number allows us to examine different worst-case contagion scenarios by varying the spread factor kk. In this paper we prove that computing kk-burning number is APX-hard, for any fixed constant kk. We then give an O((n+m)logn)O((n+m)\log n)-time 3-approximation algorithm for computing kk-burning number, for any k1k\ge 1, where nn and mm are the number of vertices and edges, respectively. Finally, we show that even if the burning sources are given as an input, computing a burning sequence itself is an NP-hard problem.

Cite

@article{arxiv.2006.14733,
  title  = {APX-Hardness and Approximation for the k-Burning Number Problem},
  author = {Debajyoti Mondal and N. Parthiban and V. Kavitha and Indra Rajasingh},
  journal= {arXiv preprint arXiv:2006.14733},
  year   = {2021}
}
R2 v1 2026-06-23T16:38:22.668Z