Update on k-truss Decomposition on GPU

Al-Masri, M.S.; Anjum, Omer; Pearson, Carl; Qureshi, Zaid; Mailthody, Vikram Sharma; Nagi, Rakesh; Xiong, Jinjun; Hwu, Wen-mei W.

doi:10.1109/hpec.2019.8916285

Cited by 21 publications

(10 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…6) MD+18 is a finalist of the 2018 GraphChallenge for k-truss decomposition, specifically designed for GPU platforms [22]. 7) AA+19 is a winner of the Student Innovation Awards in the 2019 GraphChallenge [2] To the best of our knowledge, all other known methods for k-truss decomposition in literature are directly improved or outperformed by at least one of these (see Section II for discussion).…”

Section: State-of-the-art Algorithmsmentioning

confidence: 99%

“…In particular, we will use FMT setting r = 4 and M = m/10, where m is the number of edges in the input graph G. In the following, we will refer to this method as FMT-max. We compare these with AA+19 [2], which also proposes an algorithm for max-truss computation.…”

Section: State-of-the-art Algorithmsmentioning

confidence: 99%

“…This makes a direct comparison challenging (furthermore, the software is not available). To obtain a meaningful comparison, we devised a cost-based approach that is detailed in Section VI-B, which takes into account the cost of the machines used and the time needed by that machines to conclude an experiment, using the results and dataset reported in the corresponding papers [2], [22], [25]. 8 The paper proposes a parallel algorithm too, but (as confirmed by the authors) the serial one is consistently faster even in parallel environments.…”

Section: Indirect Comparisonmentioning

confidence: 99%

“…As it can be observed, k-trusses are more rigorous than k-cores but less stringent than cliques: it can be proved that a k-truss is a subgraph of a (k − 1)core, and that a k-clique is also a k-truss. 2 Furthermore, k-trusses can be computed in polynomial time, and may also be employed to quickly remove useless edges from a graph when looking for k-cliques [10].…”

Section: Introductionmentioning

confidence: 99%

“…The trussness t G of the graph G is the maximum k such that there exists a k-truss in G, and the corresponding k-truss 2 Indeed a k-clique induces a k-truss: any two adjacent nodes in a clique of size k may form a triangle with any of the remaining k − 2 nodes. Also, as each edge of a k-truss forms at least k −2 triangles, its extremes must have degree at least k − 1 in the k-truss, meaning that a k-truss is also a k − 1-core.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Truly Scalable K-Truss and Max-Truss Algorithms for Community Detection in Graphs

et al. 2020

View full text Add to dashboard Cite

The notion of k-truss has been introduced a decade ago in social network analysis and security for community detection, as a form of cohesive subgraphs less stringent than a clique (set of pairwise linked nodes), and more selective than a k-core (induced subgraph with minimum degree k). A k-truss is an inclusion-maximal subgraph H in which each edge belongs to at least k − 2 triangles inside H . The truss decomposition establishes, for each edge e, the maximum k for which e belongs to a k-truss. Analogously to the largest clique and to the maximum k-core, the strongest community for k-truss is the max-truss, which corresponds to the k-truss having the maximum k. Even though the computation of truss decomposition and of the max-truss takes polynomial time, on a large scale, it suffers from handling a potentially cubic number of wedges. In this paper, we provide a new algorithm FMT, which advances the state of the art on different sides: lower execution time, lower memory usage, and no need for expensive hardware. We compare FMT experimentally with the most recent state-of-the-art algorithms on a set of large real-world and synthetic networks with over a billion edges. The massive improvement allows FMT to compute the max-truss of networks of tens of billions of edges on a single standard server machine.

show abstract

Section: State-of-the-art Algorithmsmentioning

confidence: 99%

Section: State-of-the-art Algorithmsmentioning

confidence: 99%