We study the parameterized and classical complexity of the s-Club Cluster Edge Deletion problem: given a graph G = (V, E) and integers k and s, determine whether it is possible to delete at most k edges so that every connected component of the resulting graph has diameter at most s. This problem generalizes Cluster Edge Deletion (the case s = 1) and captures a variety of distance-bounded graph modification tasks. Montecchiani, Ortali, Piselli, and Tappini (Information and Computation, 2023) showed that the problem is fixed-parameter tractable when parameterized by s plus the treewidth of G, and asked whether the dependence on s is necessary; that is, whether the problem is FPT when parameterized by treewidth alone. We resolve this by proving that the problem is W[1]-hard when parameterized by pathwidth, and hence by treewidth. On the algorithmic side, we show that the problem is FPT when parameterized by neighborhood diversity, twin cover, or cluster vertex deletion number, thereby extending to all s >= 1 the results of Italiano, Konstantinidis, and Papadopoulos (Algorithmica, 2023), who established FPT algorithms for the case s = 1 under the neighborhood diversity and twin cover parameters. From a classical perspective, we prove that the problem is NP-hard on split graphs already for s = 2, complementing the polynomial-time solvability for s = 1 due to Bonomo, Duran, and Valencia-Pabon (Theoretical Computer Science, 2015) and the trivial case s = 3. Finally, while the problem is FPT when parameterized by s + k, its complexity for the solution size k alone remains open. We make progress on this front by designing an FPT bicriteria approximation algorithm, which runs in time f(k, 1/epsilon) * n^{O(1)} and, for graphs excluding long induced cycles, outputs a solution of size at most k whose connected components have diameter at most (1 + epsilon) * s.

翻译：暂无翻译