2026-07-22T15:57:08Z https://drops.dagstuhl.de/oai

oai:drops-oai.dagstuhl.de:16423 2024-03-06T10:57:26Z ddc:004 open_access

Beating Matrix Multiplication for n^{1/3}-Directed Shortcuts Kogan, Shimon Parter, Merav Directed Shortcuts Transitive Closure Width For an n-vertex digraph G = (V,E) and integer parameter D, a D-shortcut is a small set H of directed edges taken from the transitive closure of G, satisfying that the diameter of G ∪ H is at most D. A recent work [Kogan and Parter, SODA 2022] presented shortcutting algorithms with improved diameter vs. size tradeoffs. Most notably, obtaining linear size D-shortcuts for D = Õ(n^{1/3}), breaking the √n-diameter barrier. These algorithms run in O(n^{ω}) time, as they are based on the computation of the transitive closure of the graph. We present a new algorithmic approach for D-shortcuts, that matches the bounds of [Kogan and Parter, SODA 2022], while running in o(n^{ω}) time for every D ≥ n^{1/3}. Our approach is based on a reduction to the min-cost max-flow problem, which can be solved in Õ(m+n^{3/2}) time due to the recent breakthrough result of [Brand et al., STOC 2021]. We also demonstrate the applicability of our techniques to computing the minimal chain covers and dipath decompositions for directed acyclic graphs. For an n-vertex m-edge digraph G = (V,E), our key results are: - An Õ(n^{1/3}⋅ m+n^{3/2})-time algorithm for computing D-shortcuts of linear size for D = Õ(n^{1/3}), and an Õ(n^{1/4}⋅ m+n^{7/4})-time algorithm for computing D-shortcuts of Õ(n^{3/4}) edges for D = Õ(n^{1/2}). - For a DAG G, we provide Õ(m+n^{3/2})-time algorithms for computing its minimum chain covers, maximum antichain, and decomposition into dipaths and independent sets. This improves considerably over the state-of-the-art bounds by [Caceres et al., SODA 2022] and [Grandoni et al., SODA 2021]. Our results also provide a new connection between shortcutting sets and the seemingly less related problems of minimum chain covers and the maximum antichains in DAGs. Schloss Dagstuhl – Leibniz-Zentrum für Informatik Shimon Kogan and Merav Parter 2022 Is Part Of LIPIcs, Volume 229, 49th International Colloquium on Automata, Languages, and Programming (ICALP 2022) InProceedings Text doc-type:ResearchArticle publishedVersion application/pdf doi:10.4230/LIPIcs.ICALP.2022.82 urn:nbn:de:0030-drops-164230 https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.ICALP.2022.82 eng https://creativecommons.org/licenses/by/4.0/legalcode