EUVIMEDObjevte EUVIMED
Open access

Higher-order trade-offs in hypergraph community detection

Jiaze Li iD, Michael T Schaub iD, Leto Peel iD

Science Advances · 2026-08-07 · 12 (32)

DOI: 10.1126/sciadv.aef2184 ↗

At a glance

Journal article in Science Advances, published on 2026-08-07. Written by Jiaze Li and 2 co-authors. Institution: Department of Data Analytics and Digitalisation, Maastricht University, Maastricht, Netherlands. 45 references. The full text is freely available in PubMed Central. Funded by: European Research Council, Nederlandse Organisatie voor Wetenschappelijk Onderzoek, China Scholarship Council.

AbstraktBibliografické údajeCiteRelated studies

Abstract (original language)

Extending community detection from pairwise networks to hypergraphs introduces fundamental theoretical challenges. Hypergraphs exhibit structural heterogeneity with no direct graph analog: Hyperedges of varying orders can connect nodes across communities in diverse configurations. Crucially, this means that any hypergraph community detection algorithm must make a choice regarding which order or balance of hyperedges it prefers to maintain or split during partitioning. We address these challenges by developing a unified framework for community detection in nonuniform hypergraphs under the hypergraph stochastic block model. We introduce a general signal-to-noise ratio that enables a quantitative analysis of trade-offs unique to higher-order networks, such as which hypergedges we choose to split across communities and how we choose to split them. Building on this framework, we derive a Bethe Hessian operator for nonuniform hypergraphs that provides efficient spectral clustering with principled model selection. We characterize the resulting spectral detectability threshold and compare it to belief propagation limits, showing the methods coincide for uniform hypergraphs but diverge in nonuniform settings. Synthetic experiments confirm our analytical predictions and reveal systematic biases toward preserving higher-order and balanced shape hyperedges. Application to empirical data demonstrates the practical relevance of these higher-order detectability trade-offs in real-world systems.

Authors and institutions

  • Department of Data Analytics and Digitalisation, Maastricht University, Maastricht, Netherlands.
  • Department of Computer Science, RWTH Aachen University, Aachen, Germany.
  • Department of Biology, RWTH Aachen University, Aachen, Germany.
  • Center for Computational Life Sciences, RWTH Aachen University, Aachen, Germany.

ORCID profiles: Jiaze Li, Michael T Schaub, Leto Peel

Funding

  • European Research Council
  • Nederlandse Organisatie voor Wetenschappelijk Onderzoek
  • China Scholarship Council

Cite

APA

Li, J., Schaub, M. T., & Peel, L. (2026). Higher-order trade-offs in hypergraph community detection. Science Advances, 12(32). https://doi.org/10.1126/sciadv.aef2184

Vancouver

Li J, Schaub MT, Peel L. Higher-order trade-offs in hypergraph community detection. Science Advances. 2026;12(32). doi:10.1126/sciadv.aef2184

BibTeX
@article{Li2026,
  author = {Li, Jiaze and Schaub, Michael T and Peel, Leto},
  title = {Higher-order trade-offs in hypergraph community detection},
  journal = {Science Advances},
  year = {2026},
  volume = {12},
  number = {32},
  doi = {10.1126/sciadv.aef2184}
}
Start for free →