Cohesive Subgraph Discovery in Hypergraphs: A Locality-Driven Indexing Framework

• DOI: 10.48550/arxiv.2502.12523
• Publication Date: 2025-02-17
• Authors: Song Lim Kim; Dahee Kim; Junghoon Kim; Hyun Ji Jeong; Jungeun Kim
• Keywords: Social and Information Networks (cs.SI); FOS: Computer and information sciences

Hypergraphs are increasingly employed to model complex, diverse relationships in modern networks, effectively capturing higher-order interactions. A critical challenge this domain is the discovery of cohesive subgraphs, which provides valuable insights into hypergraph structures. However, selecting suitable parameters for task remains unresolved. To address this, we propose an efficient indexing framework designed online retrieval subgraphs. Our approach enables rapid identification desired structures without requiring exhaustive graph traversals, thus ensuring scalability and practicality. This has broad applicability, supporting informed decision-making across various domains by offering a comprehensive view network landscapes. Extensive experiments on real-world datasets demonstrate effectiveness efficiency our proposed technique.