The simpliciality of higher-order networks

EPJ Data Science, Год журнала: 2024, Номер 13(1)

Опубликована: Март 7, 2024

Higher-order networks are widely used to describe complex systems in which interactions can involve more than two entities at once. In this paper, we focus on inclusion within higher-order networks, referring situations where specific participate an interaction, and subsets of those also interact with each other. Traditional modeling approaches tend either not consider all (e.g., hypergraph models) or explicitly assume perfect complete simplicial models). To allow for a nuanced assessment introduce the concept "simpliciality" several corresponding measures. Contrary current practice, show that empirically observed rarely lie end simpliciality spectrum. addition, generative models fitted these datasets struggle capture their structure. These findings suggest new directions field network science.

Язык: Английский

---

Hyperlink communities in higher-order networks

Journal of Complex Networks, Год журнала: 2024, Номер 12(2)

Опубликована: Фев. 21, 2024

Abstract Many networks can be characterized by the presence of communities, which are groups units that closely linked. Identifying these communities crucial for understanding system’s overall function. Recently, hypergraphs have emerged as a fundamental tool modelling systems where interactions not limited to pairs but may involve an arbitrary number nodes. In this study, we adopt dual approach community detection and extend concept link hypergraphs. This extension allows us extract informative clusters highly related hyperedges. We analyse dendrograms obtained applying hierarchical clustering distance matrices among hyperedges across variety real-world data, showing hyperlink naturally highlight multiscale structure higher-order networks. Moreover, enable overlapping memberships from nodes, overcoming limitations traditional hard methods. Finally, introduce network cartography practical categorizing nodes into different structural roles based on their interaction patterns participation. aids in identifying types individuals social systems. Our work contributes better organization

Язык: Английский

---

Framework to generate hypergraphs with community structure

Physical review. E, Год журнала: 2024, Номер 109(3)

Опубликована: Март 19, 2024

In recent years hypergraphs have emerged as a powerful tool to study systems with multibody interactions which cannot be trivially reduced pairs. While highly structured methods generate synthetic data proved fundamental for the standardized evaluation of algorithms and statistical real-world networked data, these are scarcely available in context hypergraphs. Here we propose flexible efficient framework generation many nodes large hyperedges, allows specifying general community structures tune different local statistics. We illustrate how use our model sample desired features (assortative or disassortative communities, mixed hard assignments, etc.), analyze detection algorithms, structurally similar data. Overcoming previous limitations on hypergraphs, work constitutes substantial advancement modeling higher-order systems.

Язык: Английский

---

Inference and visualization of community structure in attributed hypergraphs using mixed-membership stochastic block models

Social Network Analysis and Mining, Год журнала: 2025, Номер 15(1)

Опубликована: Март 10, 2025

Язык: Английский

---

Estimating Hyperedge Size Distribution via Random Walk on Hypergraphs

Studies in computational intelligence, Год журнала: 2025, Номер unknown, С. 68 - 76

Опубликована: Янв. 1, 2025

Язык: Английский

---

Patterns in Temporal Networks with Higher-Order Egocentric Structures

Entropy, Год журнала: 2024, Номер 26(3), С. 256 - 256

Опубликована: Март 13, 2024

The analysis of complex and time-evolving interactions, such as those within social dynamics, represents a current challenge in the science systems. Temporal networks stand suitable tool for schematizing systems, encoding all interactions appearing between pairs individuals discrete time. Over years, network has developed many measures to analyze compare temporal networks. Some them imply decomposition into small pieces interactions; i.e., only involving few nodes short time range. Along this line, possible way decompose is assume an egocentric perspective; consider each node evolution its neighborhood. This was proposed by Longa et al. defining “egocentric neighborhood”, which proven be useful characterizing relative interactions. However, definition neglects group (quite common domains), they are always decomposed pairwise connections. A more general framework that also allows considering larger represented higher-order Here, we generalize description hypergraphs. Consequently, their “hyper neighborhoods”. enables facilitating comparisons different datasets or dataset, while intrinsic complexity presented Even if limit order second (triplets nodes), our results reveal importance representation.In fact, analyses show second-order structures responsible majority variability at scales: datasets, amongst nodes, over

Язык: Английский

---

Comparison of modularity-based approaches for nodes clustering in hypergraphs

Peer Community Journal, Год журнала: 2024, Номер 4

Опубликована: Март 22, 2024

Statistical analysis and node clustering in hypergraphs constitute an emerging topic suffering from a lack of standardization. In contrast to the case graphs, concept nodes' community is not unique encompasses various distinct situations. this work, we conducted comparative performance modularity-based methods for nodes binary hypergraphs. To address this, begin by presenting, within unified framework, hypergraph modularity criteria proposed literature, emphasizing their differences respective focuses. Subsequently, provide overview state-of-the-art codes available maximize modularities detecting communities Through exploration simulation settings with controlled ground truth clustering, offer comparison these using different quality measures, including true recovery, running time, (local) maximization objective, number clusters detected. Our contribution marks first attempt clarify advantages drawbacks newly methods. This effort lays foundation better understanding primary objectives

Язык: Английский

---

HEDV-Greedy: An Advanced Algorithm for Influence Maximization in Hypergraphs

Mathematics, Год журнала: 2024, Номер 12(7), С. 1041 - 1041

Опубликована: Март 30, 2024

Influence maximization (IM) has shown wide applicability in various fields over the past few decades, e.g., viral marketing, rumor control, and prevention of infectious diseases. Nevertheless, existing research on IM primarily focuses ordinary networks with pairwise connections between nodes, which fall short representation higher-order relations. hypergraphs (HIM) received limited attention. A novel evaluation function, aims to evaluate spreading influence selected nodes hypergraphs, i.e., expected diffusion value hypergraph (HEDV), is proposed this work. Then, an advanced greedy-based algorithm, termed HEDV-greedy, select seed maximum hypergraph. We conduct extensive experiments eight real-world datasets, benchmarking HEDV-greedy against state-of-the-art methods for HIM problem. Extensive conducted datasets highlight effectiveness efficiency our methods. The algorithm demonstrates a marked reduction time complexity by two orders magnitude compared conventional greedy method. Moreover, outperforms other algorithms across all datasets. Specifically, under conditions lower propagation probability, exhibits average improvement solution accuracy 25.76%.

Язык: Английский

---

Developing the ‘omic toolkit of comparative physiologists

Comparative Biochemistry and Physiology Part D Genomics and Proteomics, Год журнала: 2024, Номер 52, С. 101287 - 101287

Опубликована: Июль 3, 2024

Typical 'omic analyses reduce complex biological systems to simple lists of supposedly independent variables, failing account for changes in the wider transcriptional landscape. In this commentary, we discuss utility network approaches incorporating context into study physiological phenomena. We highlight opportunities build on traditional tools by utilising cutting-edge techniques higher order interactions (i.e. beyond pairwise associations) within datasets, allowing more accurate models systems. Finally, show examples previous works gain additional insight their organisms interest. As 'omics grow both popularity and breadth application, so does requirement flexible analytical capable interpreting synthesising datasets.

Язык: Английский

---

Optimal control of spatial diseases spreading in networked reaction–diffusion systems

Physics Reports, Год журнала: 2025, Номер 1111, С. 1 - 64

Опубликована: Фев. 1, 2025

Язык: Английский

---