CS Paper Radar

Sparsification Framework for Directed Densest Subgraph

Slobodan Mitrović, Theodore Pan

cs.DS ·submitted 2026-10-01 00:06 UTC · v1

Abstract

We develop a new approach for computing approximate directed densest subgraphs (DDS). Our main result is a sparsification procedure that reduces a directed graph $G$ on $n$ vertices to a graph with $n \cdot \text{poly} \log n$ edges while preserving enough structure to recover an approximate DDS of $G$. Instantiating this framework in several memory-constrained settings, we obtain the following improvements over the state of the art: In semi-streaming, we obtain a single-pass algorithm that computes a $(1-\varepsilon)$-approximate DDS. Previously, the only semi-streaming algorithm that computed a constant approximation of DDS was by Bahmani, Kumar, and Vassilvitskii (2012), providing a $0.5-\varepsilon$ approximation in $O(\log n)$ passes. Hence, our work completely closes the approximation gap between undirected and directed DS in the semi-streaming setting, matching the $(1-\varepsilon)$-approximate undirected DS algorithm by Esfandiari, Hajiaghayi, and Woodruff (2016). In the near-linear-memory MPC regime, we obtain an $O(1)$-round algorithm for $(1-\varepsilon)$-approximate DDS, improving over the $O(\sqrt{\log n})$-round $(0.5-\varepsilon)$-approximation algorithm of Mitrović and Pan (2024). In the sublinear-time setting, we obtain an algorithm using $\tilde{O}(n)$ time, space, and oracle queries to compute a $(1-\varepsilon)$-approximate DDS, improving over the $\tilde{O}(n^{1.5})$ time, space, and query algorithm of Esfandiari, Hajiaghayi, and Woodruff (2016).

Closest papers

Nearest neighbours by meaning across the whole index.

Semi-Streaming Reachability and Shortest Paths in $n^{1/3+o(1)}$ Passes

cs.DS ·2026-11-03 ·Gary Hoppenworth, George Z. Li, Sepideh Mahabadi, Jakub Tarnawski
theory similarity 0.72
Abstract

In the semi-streaming model, the edges of an $n$-vertex graph arrive in a stream, and an algorithm makes multiple passes over that stream while using only $\widetilde{O}(n)$ space. The number of passes needed to compute single-source reachability in directed graphs is a long-standing open question in streaming algorithms, going back to Henzinger, Raghavan, and Rajagopalan (1998). The best known semi-streaming algorithm uses $n^{1/2+o(1)}$ passes due to Jambulapati, Liu, and Sidford (2019), while the best lower bound is $Ω(\log n/\log\log n)$ passes due to Guruswami and Onak (2016). We give streaming algorithms for computing single-source reachability and shortest paths on $n$-vertex directed graphs using $\widetilde{O}(n)$ space and $n^{1/3+o(1)}$ passes. Our algorithms are based on implicit hopsets, which are large hopsets that can be represented implicitly in $\widetilde{O}(n)$ space. We introduce this notion and give new constructions for implicit hopsets with small hopbound. Along the way, we obtain new work-depth tradeoffs for parallel shortest paths in directed graphs, improving the recent result of Ashvinkumar, Bernstein, Probst Gutenberg, and Saranurak (2026).

0.20

A Linear-Time Approximation Scheme for the Densest Subgraph Problem

cs.DS ·2026-08-11 ·Elena Grigorescu, Mehrshad Taziki
similarity 0.71
Abstract

In the undirected \emph{Densest Subgraph Problem (DSG)} the goal is to output a subset $S$ of vertices of a given graph $G$ that maximizes the quantity $|E(S)|/|S|$, where $E(S)$ is the set of edges in the subgraph induced by $S$. The problem is well studied in both theory and practice, and it admits natural efficient exact algorithms, as well as near-linear time algorithms with a $(1-\varepsilon)$ approximation ratio. However, all previously-known approximation schemes incur logarithmic factors in the size of the graph or other parameters of the graph. This raises the question of whether a linear time $(1-\varepsilon)$-approximation can be obtained for all $\varepsilon>0$. We answer this question affirmatively by providing a $(1-\varepsilon)$-approximation algorithm running in time $O\left(\frac{n+m}{\varepsilon^3}\log \frac{1}{\varepsilon}\right)$, where $m$ and $n$ are respectively the number of edges and vertices of $G$. To the best of our knowledge, this is the first truly linear-time approximation scheme for the problem (when $\varepsilon>0$ is a constant). Our algorithm uses assignments arising from a flow-based formulation together with a structural carving lemma. This lemma allows us to progressively carve "sparse" parts of the graph while nearly preserving the densest subgraph, allowing us to shift heavy computations to smaller instances, which eventually yields the mentioned runtime. Our framework also yields a $(1/2 -\varepsilon)$-approximation for the \emph{Densest At-Least-$k$ Subgraph Problem}, where in addition to maximizing the density, we require the subgraph to have at least $k$ vertices. Our algorithm runs in time $O\left( \frac{(n+m) \log^2 n \log \frac{1}{\varepsilon}}{\varepsilon} \right)$. This nearly matches the known $1/2$ approximation hardness while running in near-linear time.

0.04

A Nearly Optimal Min-Cut Algorithm for Dynamic Streams

cs.DS ·2026-10-25 ·Yang P. Liu, Hoai-An Nguyen, Vihan Shah, David P. Woodruff
similarity 0.70
Abstract

We give an $\tilde{O}(n/\varepsilon)$-space algorithm for the minimum cut problem in dynamic streams, resolving the space complexity of streaming min-cut up to polylogarithmic factors. Ding, Garces, Li, Lin, Nelson, Shah, and Woodruff (ITCS 2025) obtained the same $\tilde{O}(n/\varepsilon)$ space bound for insertion-only streams, while prior to our work the best known bound even for unweighted dynamic streams was $\tilde{O}(n/\varepsilon^2)$, due to Ahn, Guha, and McGregor (PODS 2012). Our main technical contribution is an $\tilde{O}(n/\varepsilon)$-space for-each cut sketch in dynamic streams, which may be of independent interest. The construction applies to weighted graphs under general $\pmΔ$ updates, a stronger model than the standard dynamic-stream setting in which weighted edges are inserted and deleted in their entirety. Despite this stronger guarantee, our algorithm is surprisingly simple and relies on very little machinery. We also obtain applications to minimum $\ell_p$-norm $k$-partitioning for every fixed $k$, weighted correlation-clustering evaluation for two clusters, and for-each cut sketching in balanced directed graphs.

0.46

Square-Root Pass Semi-Streaming Algorithms for Digraph Problems

cs.DS ·2026-11-02 ·Prantar Ghosh, Amirhossein Janfaza
theory similarity 0.69
Abstract

We study directed graph problems in the semi-streaming model where we are restricted to $\widetilde{O}(n):=O(n\cdot \text{polylog}(n))$ space for processing the edge stream of an $n$-node graph and can make one or a few passes to solve the given problem. A longstanding open problem in multipass streaming is determining the number of semi-streaming passes needed to solve directed $s,t$-reachability: there is an exponential gap---$Ω(\log n/\log \log n)$ vs $n^{1/2+o(1)}$---between the best-known lower and upper bounds. Several related fundamental digraph problems, e.g., topological sorting, acyclicity testing, and decomposition into strongly connected components (SCCs), are known to have the same semi-streaming pass lower bound. However, no $o(n/\text{polylog}(n))$-pass upper bound was known for any of them. Our main result achieves an $n^{1/2+o(1)}$-pass semi-streaming algorithm for each of these problems. As a corollary, we get algorithms with the same pass and space usage for a broad class of DAG problems that admit a certain form of dynamic program over a topological ordering. This class includes weighted single-source shortest paths, longest paths, max-outdegree+1 coloring, and constructing a lexicographically-first DFS (depth-first search) forest. Our algorithm for the last problem can be seen as a generalization of our results on topological sorting and acyclicity testing: recall that standard textbook algorithms for these problems use DFS as a subroutine. Our algorithms are optimal up to an $n^{o(1)}$ factor in passes under the reasonable conjecture that $s,t$-reachability requires $Ω(\sqrt{n})$ semi-streaming passes. However, using the fact that our algorithms can be parameterized by the number of passes taken by topological sorting or single-source reachability algorithms, we prove stronger lower bounds for some of the problems even under weaker conjectures.

0.28

Improved Pass Complexity for Semi-Streaming Matching in General Graphs via Budgeted Price Approach

cs.DS ·2026-10-15 ·Tatsuya Terao
similarity 0.69
Abstract · 11 pages

We present a simple deterministic semi-streaming algorithm that computes a $(1-\varepsilon)$-approximate maximum matching in general graphs in $O(1/\varepsilon^{2})$ passes. This matches the pass complexity of the auction algorithm of Assadi, Liu, and Tarjan~[SOSA'21] for bipartite graphs, and answers Open Question~2 of Assadi~[SOSA'24 \& TheoretiCS'25]. Our approach, which we call the \emph{budgeted price approach}, is very different from the previous $\mathrm{poly}(1/\varepsilon)$-pass semi-streaming algorithms for $(1-\varepsilon)$-approximation of matching in general graphs and improves their pass complexity. All previous approaches are augmenting-path-based and use $O(1/\varepsilon^{19})$ passes by Fischer, Mitrović, and Uitto~[STOC'22], $O(1/\varepsilon^{6})$ passes by Mitrović, Mukherjee, Sankowski, and Sheu~[ICALP'25], and $O(1/\varepsilon^{4})$ passes by Izumi, Kitamura, and Yamaguchi~['25].

0.26

Near-Optimal Space and Time Cut Sparsification in Turnstile Streams

cs.DS ·2026-11-02 ·Honghao Lin, Vahab Mirrokni, David P. Woodruff
similarity 0.68
Abstract

A seminal result of Ahn, Guha, and McGregor (PODS'12) showed that a cut sparsifier of an unweighted graph can be computed from a near-linear number of linear measurements. Later work extended this to spectral sparsifiers (Kapralov et al., FOCS'14) and to near-linear recovery time (Kapralov et al., SODA'20). For weighted graphs, these sketches rely on grouping edges by weight, which general turnstile streams do not allow. Chen, Khanna, and Li (FOCS'22) removed this limitation, using $\tilde{O}(n/\varepsilon^3)$ bits of space. We improve the space to $\tilde{O}(n/\varepsilon^2)$ bits. Our algorithm makes one pass over a general turnstile stream of a weighted graph, processes each update in $\tilde{O}(1)$ time, and recovers, in $\tilde{O}(n/\varepsilon^2)$ time, a reweighted subgraph with $O(n\varepsilon^{-2}\log^2 n)$ edges that preserves every cut within a factor of $1\pm\varepsilon$. Its space, update time, and recovery time are all optimal up to polylogarithmic factors.

0.24

Combinatorial Minimum Cost Flow in Almost-Linear Time on Dense Graphs

cs.DS ·2026-10-23 ·Bernhard Haeupler, Yonggang Jiang, Thatchaphol Saranurak
similarity 0.66
Abstract · 101 pages

We give a combinatorial algorithm that computes an exact minimum-cost flow in $n^{2+o(1)}$ time. This is the first combinatorial improvement since Goldberg-Tarjan [GT87], and is almost optimal in dense graphs. All previous faster algorithms, including the almost-linear-time algorithm of Chen et al. [CKL+22], rely on interior-point methods. Our main conceptual contribution is a black-box reduction from exact minimum-cost flow on general graphs to its approximation on DAGs, with only an $n^{o(1)}$ loss on the running time. This reduction allows future work on combinatorial minimum-cost flow to focus on DAGs, typically much easier to handle than general graphs. Our $n^{2+o(1)}$-time algorithm follows by combining this reduction with a simple augmenting-path-based approximation algorithm on DAGs. The reduction is based on a new structural result, called a flow-cost-preserving DAG projection. Given an arbitrary directed graph, this construction produces a DAG that approximately preserves both the congestion of flows and their cost up to a small loss.

0.49

Query Complexity of Densest Sub(hyper)graph Problems

cs.DS cs.CC ·2026-10-13 ·Sinan G. Aksoy, Eli Friedman, Nate Veldt, Stephen J. Young
theory similarity 0.66
Abstract · 26 pages

Rubinstein et al. introduce the study of connectivity and minimum cut problems in the cut query model, where an algorithm can only access the edges of a graph through CUT queries. In this work, we examine the query complexity of densest subgraph problems. Our first contribution is to generalize the uniform sparsification approach for Densest Subgraph and other heavy subgraph problems to weighted graphs, achieving the following results. >A $(1+\varepsilon)$-approximation for weighted Densest Subgraph in $O(n\log^2 n/\varepsilon^2)$ queries, roughly matching the $Ω(n/\log n)$ lower bound we prove for randomized algorithms. >A $(1+\varepsilon)$-approximation for weighted Max Cut in $O(n\log^2 n/\varepsilon^2)$ queries. This improves upon the bound of $O(n\log^3 n(\log n+1/\varepsilon^2))$ queries achieved by Plevrakis et al., making significant progress towards their lower bound of $Ω(n/\log n)$ queries. In contrast, we rule out approximation schemes with truly subquadratic query complexity for Densest $k$-Subgraph and settle its exact query complexity in both the unweighted and deterministic settings. For hypergraphs, we show CUT queries alone cannot achieve $o(\sqrt{n})$-approximation for Densest Subhypergraph or $o(n)$-approximation for Densest $k$-Subhypergraph, regardless of the number of queries. Augmenting CUT queries with the EDGE oracle of Chen et al., we overcome this barrier for Densest Subhypergraph and obtain a $(1+\varepsilon)$-approximation on rank-$r$ hypergraphs using $O(nr\log n/\varepsilon^2)$ queries. We also extend our Densest $k$-Subgraph hardness results to hypergraphs and prove exponential query lower bounds for exact Densest $k$-Subhypergraph. Lastly, confirming a conjecture of Chakrabarty and Liao, we show that simulating an ADD query using CUT and IS-EDGE$^?$ queries requires exponentially-many queries.

0.24