Neighborhood-based Hypergraph Core Decomposition

被引:8
作者
Arafat, Naheed Anjum [1 ]
Khan, Arijit [2 ]
Rai, Arpit Kumar [3 ]
Ghosh, Bishwamittra [1 ]
机构
[1] NUS, Singapore, Singapore
[2] AAU, Aalborg, Denmark
[3] IIT Kanpur, Kanpur, Uttar Pradesh, India
来源
PROCEEDINGS OF THE VLDB ENDOWMENT | 2023年 / 16卷 / 09期
基金
新加坡国家研究基金会;
关键词
ALGORITHMS; NETWORKS;
D O I
10.14778/3598581.3598582
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
We propose neighborhood-based core decomposition: a novel way of decomposing hypergraphs into hierarchical neighborhood-cohesive subhypergraphs. Alternative approaches to decomposing hypergraphs, e.g., reduction to clique or bipartite graphs, are not meaningful in certain applications, the later also results in inefficient decomposition; while existing degree-based hypergraph decomposition does not distinguish nodes with different neighborhood sizes. Our case studies show that the proposed decomposition is more effective than degree and clique graph-based decompositions in disease intervention and in extracting provably approximate and application-wise meaningful densest subhypergraphs. We propose three algorithms: Peel, its efficient variant E-Peel, and a novel local algorithm: Local-core with parallel implementation. Our most efficient parallel algorithm Local-core(P) decomposes hypergraph with 27M nodes and 17M hyperedges in-memory within 91 seconds by adopting various optimizations. Finally, we develop a new hypergraph-core model, the (neighborhood, degree)-core by considering both neighborhood and degree constraints, design its decomposition algorithm Local-core+Peel, and demonstrate its superiority in spreading diffusion.
引用
收藏
页码:2061 / 2074
页数:14
相关论文
共 57 条
[1]  
Alvarez-Hamelin J., 2005, ADV NEURAL INFORM PR
[2]  
Arafat NA, 2023, Arxiv, DOI arXiv:2301.06426
[3]   Construction and Random Generation of Hypergraphs with Prescribed Degree and Dimension Sequences [J].
Arafat, Naheed Anjum ;
Basu, Debabrota ;
Decreusefond, Laurent ;
Bressan, Stephane .
DATABASE AND EXPERT SYSTEMS APPLICATIONS, DEXA 2020, PT II, 2020, 12392 :130-145
[4]  
Arafat Naheed Anjum, 2022, OUR CODE DATASETS
[5]   Random Preferential Attachment Hypergraph [J].
Avin, Chen ;
Lotker, Zvi ;
Nahum, Yinon ;
Peleg, David .
PROCEEDINGS OF THE 2019 IEEE/ACM INTERNATIONAL CONFERENCE ON ADVANCES IN SOCIAL NETWORKS ANALYSIS AND MINING (ASONAM 2019), 2019, :398-405
[6]  
Bahmanian, 2015, THEORY APPL GRAPHS, V2
[7]  
Bailey Stephen, 2017, NASHVILLE MEETUP NET
[8]  
Batagelj V, 1999, LECT NOTES COMPUT SC, V1731, P90
[9]   Fast algorithms for determining (generalized) core groups in social networks [J].
Batagelj, Vladimir ;
Zaversnik, Matjaz .
ADVANCES IN DATA ANALYSIS AND CLASSIFICATION, 2011, 5 (02) :129-145
[10]   Networks beyond pairwise interactions: Structure and dynamics [J].
Battiston, Federico ;
Cencetti, Giulia ;
Iacopini, Iacopo ;
Latora, Vito ;
Lucas, Maxime ;
Patania, Alice ;
Young, Jean-Gabriel ;
Petri, Giovanni .
PHYSICS REPORTS-REVIEW SECTION OF PHYSICS LETTERS, 2020, 874 :1-92