Pith. sign in

REVIEW 3 major objections 3 minor 49 references

The recurrence of groups inhibits the information spreading under higher-order interactions

T0 review · 3 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read In face-to-face interaction networks, the recurrence of triangular groups inhibits the spread of information, and the effect grows stronger when higher-order group interactions are present.

desk verdict Recurrent triangles are real and the inhibition effect is plausible, but the null model conflates edge and triangle recurrence, so the title claim needs a control. read the letter →

arxiv 2502.09348 v1 pith:CLKMEOA4 submitted 2025-02-13 physics.soc-ph nlin.CD

Liang Yuan , Jiao Wu , Kesheng Xu , Muhua Zheng This is my paper · ORCID
classification physics.soc-phnlin.CD
keywords recurrenceofgroupsgroupformationinformationspreadinghigher-orderinteractionsface-to-faceinteractionnetworkstemporalnullmodelSIS
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks what recurring group meetings do to the spread of information in real face-to-face interaction networks. Analysing four datasets from a primary school, a high school, a hospital ward, and a conference, the authors find that the same three-person groups reappear across time windows. They extend a force-directed motion model, in which people move toward similar others in a hidden space, and show it reproduces these recurrent triangular groups and, unlike a baseline attractiveness model, predicts how far an infection or piece of information spreads. By comparing each real network with a null model that keeps the same number of contacts and triangles in every time window but destroys their recurrence, they establish that recurring triangles suppress spreading, and that including three-way (higher-order) infection channels makes the suppression more pronounced.

What carries the argument

The load-bearing object is the null model: in each aggregated snapshot, the labels of interacting nodes are randomly swapped, so every snapshot retains its exact number of edges and triangles but the same triangle rarely reappears. Comparing the higher-order SIS spreading dynamics, with a pairwise infection rate and a triangular infection rate, on the real snapshots versus the null snapshots isolates the effect of group recurrence; a microscopic Markov chain approach gives a theoretical check on the simulations.

What would settle it

Build a null model that reconnects edges so that each pair of people meets with the same frequency as in the real data but the same three-way triangles rarely close; if the spreading range rises back to the real-network level, then pair-level recurrence is the suppressing mechanism and the triangle-level claim is not needed.

Watch

Extended reading notes

Core claim

The central claim is that recurrence of triangular groups—the same three individuals forming a full triangle in multiple aggregated time windows—reduces the final range of an SIS-type spreading process on temporal face-to-face networks. The paper demonstrates this by constructing a null model that randomly permutes the labels of interacting nodes within each snapshot, preserving the number of edges and triangles per snapshot while destroying their temporal recurrence. On all four real datasets, and in networks generated by the extended FDM model, the real or recurrent networks yield a smaller final infection density than the null counterparts; the gap widens when infection can spread through full triangles. The authors interpret this as information being trapped inside repeatedly co-occurring groups, with higher-order interactions deepening the localization.

Load-bearing premise

The null model shuffles the labels of all interacting people in each time window, keeping the number of contacts and triangles the same but wiping out every temporal memory; the claim that recurring triangles—rather than simply repeated pairs or persistent activity—cause the slowdown depends on that shuffling not also removing another factor that matters.

Editorial extensions

If this is right

  • If group recurrence is as inhibitory as the paper argues, then measures that break up recurring conversation circles—rotating seating, mixing teams—should increase the reach of information without adding new contacts.
  • Epidemic and rumor models of face-to-face contact should include triangle-level temporal memory; omitting it will overestimate the final outbreak size.
  • The gap between real and null networks grows with the higher-order infection rate, so interventions that reduce three-way contact are disproportionately effective at containing spread.
  • The FDM model, once calibrated on recurrence statistics, can serve as a generator of realistic temporal networks on which to pre-test containment strategies.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A stricter test would rewire edges to break only three-way closures while preserving each pair's repeat-contact pattern; the authors do not run this test, and the title claim could overstate the role of triangles if edge recurrence alone reproduces the suppression.
  • The localization mechanism suggests a link to echo chambers: if the same small group repeatedly interacts, information diversity inside that group falls, a testable analogue in social media would measure retweet-group recurrence against reach.
  • The result implies a network-design principle: information campaigns on campus or workplace networks should target rare, non-recurrent contacts such as newcomers or visitors rather than the dense recurring triangles.
  • A direct extension would vary triangle-recurrence level systematically in the FDM model, for instance by tuning the similarity-force strength, and map the final spreading range against recurrence rate to confirm a monotone relation without relying on a null model.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 3 minor

Summary. This manuscript studies recurrent triangular groups in four face-to-face interaction datasets, proposes an extension of the force-directed motion (FDM) model with similarity forces to reproduce the observed recurrence, and uses a higher-order SIS spreading model to argue that recurrence of triangular groups inhibits information spreading, especially under higher-order interactions. The empirical analysis is backed by an MMCA theoretical treatment, and the authors compare real data with FDM and attractiveness model (AM) simulations and with a null model obtained by random label swapping in each snapshot.

Significance. If the central claim is established, the paper makes a valuable contribution by connecting group-level temporal recurrence to spreading dynamics and by providing a generative model that reproduces several empirical properties. The work uses multiple real-world datasets, includes a theoretical MMCA analysis, and makes code and data available. The main strength is the systematic comparison of real, modeled, and null networks. However, the key causal attribution in Section 5.4 depends on a null model that removes all temporal correlations, not only triangle recurrence, so the specific claim that triangle recurrence is the operative mechanism needs additional support.

major comments (3)
  1. [Section 5.4 and SI Section II] The null model randomly permutes node labels within each 10-minute snapshot. This procedure preserves the per-snapshot number of edges and triangles but destroys every cross-snapshot correlation, including edge recurrence, node activity persistence, and duration memory. The paper interprets the decrease in final spreading range rho* as evidence that recurrence of triangular groups inhibits spreading. However, because a recurrent triangle is built from recurrent edges, the null manipulation removes edge-level recurrence at the same time. The observed effect could be driven by repeated pairwise contacts rather than by the recurrence of triangular groups specifically. To support the title claim, the authors should use a control null model that preserves edge recurrence (e.g., by shuffling triangle identities conditional on the repeated edges, or by rewiring triangles while keeping the edge set of each snapshot) or otherwise show that edge-recurrence-only effects cannot explain the gap. Without such a control, the central causal statement in the abstract and conclusion is not established.
  2. [SI Section I A and Section 5.2] The FDM parameters F0 and mu2 are tuned to match the total number of recurrent groups c of the real network, in addition to matching n and l. The main text states that the FDM model 'reproduces' the recurrent group patterns, but this reproduction is partly by construction because the model is fitted to c. This does not invalidate the null-model comparison in Section 5.4, which uses real data, but it weakens the mechanistic claim in Section 5.2 that the FDM model independently predicts group recurrence. The authors should explicitly state in the main text that c is a fitted target, and they should clarify which aspects of the recurrent group patterns (e.g., inter-event time distributions, as in Fig. 2) are genuinely predicted rather than fitted.
  3. [Section 5.4 and Fig. 5] The paper does not report error bars or confidence intervals for the empirical-versus-null difference in rho*. Given that the null model is stochastic, the authors should quantify the variance across null realizations (even if only a small number are used) and assess whether the observed suppression is statistically significant for each dataset and for beta_delta = 0 versus 0.3. This would strengthen the conclusion that the effect is robust and not driven by a particular realization or by a single dataset.
minor comments (3)
  1. [Section 4.2, Eq. (6)] The MMCA equation assumes independence between the states of neighbors at each time step, which is a standard approximation; stating this assumption explicitly would help readers who are not specialists in temporal-network MMCA.
  2. [Section 3.1, Eq. (2)-(3)] The notation in the motion equations is slightly unclear: the denominator is written as the square root of (X_j^t - X_i^t)^2 + (Y_j^t - Y_i^t)^2, but the numerator also contains the difference; this is likely a typesetting issue. Clarify the vector form.
  3. [Figure 1 caption] The caption states that purple lines represent recurrent full triangles and black lines correspond to the first occurrence, but the label 'recurrent' is defined in the main text only later; define it directly in the caption for clarity.

Circularity Check

1 steps flagged · score 2.0 of 10

Only minor partial circularity: F0 and mu2 are fitted to the recurrent-group count before the same recurrence is reported as a model reproduction; the main null-model finding is independent.

  1. fitted input called prediction [SI Section I A (Parameter tuning for FDM model); main text Section 5.2]
    "The parameters F0 and µ2 are tuned to match the average number of recurrent groups over 10-minute intervals in the simulation to the real dataset ... we filter the parameters based on the average group count c. We keep the parameter sets if |csimulated−creal|/creal < 0.2."

    Section 5.2 presents the FDM model's recurrent full-triangle patterns (Figs. 1e-h) as evidence that similarity forces explain group recurrence. But the model parameters F0 and µ2 were explicitly tuned so that the simulated average number of recurrent groups matches the real c within 20%. The 'reproduction' of recurrent groups is therefore partly enforced by the fitting target rather than independently predicted. The circularity is confined to this model-validation claim; the Section 5.4 null-model comparison does not use any fitted parameter.

full rationale

The paper's central mechanistic claim that recurrent triangular groups inhibit information spreading is tested by comparing real temporal snapshots to a null model that randomly relabels interacting nodes per snapshot, preserving per-window edge and triangle counts. This comparison involves no fitted parameters, so the suppression result is not circular. The only partial circularity is in Section 5.2 / SI Section I A: the FDM parameters F0 and mu2 are selected to match the real recurrent-group count c, and the model is then shown to reproduce recurrent-group patterns. This is a fitted input presented as reproduction, though the temporal pattern details and triangle duration/interval distributions are additional, non-fitted outputs. The null-model design does destroy edge recurrence as well as triangle recurrence, which is a potential confound for the title mechanism, but a confound is a validity concern rather than circularity under the rules of this pass. Self-citations (e.g., Ref. [14]) supply the spreading model but are not load-bearing uniqueness claims.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central results rest on the FDM model's latent similarity space, the higher-order SIS contagion model, the MMCA mean-field approximation, and, most importantly, the label-swap null model whose interpretive validity is the paper's weakest point. Four FDM parameters per dataset are fitted to the estimation portion of the empirical data, including the recurrent-group count c, so the model comparison is not a parameter-free test.

free parameters (6)
  • L (Euclidean space side length), FDM = Primary School 62, High School 94, Hospital 128, Conference 177
    Tuned so that the average degree of the time-aggregated synthetic network matches the real network (SI Section I.A).
  • μ1 (decay constant for interaction duration), FDM = 0.85, 2.10, 0.68, 2.10
    Tuned so that the average contact duration in simulation matches the real dataset (SI Section I.A).
  • F0 (force magnitude), FDM = 0.13, 0.40, 0.10, 0.04
    Tuned jointly with μ2 to match the average number of recurrent groups c and the maximum group size in 10-minute windows (SI Section I.A).
  • μ2 (force decay constant), FDM = 0.82, 0.15, 1.12, 1.27
    Tuned jointly with F0 to match c and group-size statistics (SI Section I.A).
  • Twarmup (warmup time slots), FDM = Primary School 2000, High School 6500, Hospital 2500, Conference 6000
    Chosen so that the average number of interacting individuals per slot stabilizes (SI Section I.A, Table S1).
  • L (Euclidean space side length), AM = 48, 76, 38, 78
    Tuned to match average n and l within error 0.2 (SI Section I.B).
assumptions (4)
  • domain assumption Agents live in a hidden one-dimensional similarity space; pairwise similarity distances generate forces that drive physical motion (FDM model, Section 3.1)
    Adopted from Ref. [11]; it is the mechanism used to explain recurrent groups, but it is not independently tested in this paper.
  • domain assumption The higher-order SIS model with pairwise rate β and triangle rate βΔ is an appropriate model of information spreading (Section 4.1)
    Adopted from Ref. [12]; information propagation is assumed to follow the same rules as a disease with a triangle-level transmission term.
  • domain assumption MMCA independence approximation applies to temporal networks (Section 4.2, Eqs. 6-8)
    The equations omit correlations between node states and treat neighbors' infection probabilities as independent; standard in the literature but not derived here.
  • ad hoc to paper The label-swap null model isolates triangle recurrence while preserving per-snapshot structure (SI Section II)
    This is the load-bearing assumption for the main causal claim; it is not established that only triangle recurrence is removed, since edge recurrence is also destroyed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The recurrence of groups inhibits the information spreading under higher-order interactions." pith.science (2026). https://pith.science/paper/CLKMEOA4

@misc{pith2026250209348,
  author       = {Pith},
  title        = {Pith review of: The recurrence of groups inhibits the information spreading under higher-order interactions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CLKMEOA4}},
  note         = {Machine review of arXiv:2502.09348}
}
read the original abstract

Modeling social systems as networks based on pairwise interactions between individuals offers valuable insights into the mechanisms underlying their dynamics. However, the majority of social interactions occur within groups of individuals, characterized by higher-order structures. The mechanisms driving group formation and the impact of higher-order interactions, which arise from group dynamics, on information spreading in face-to-face interaction networks remain insufficiently understood. In this study, we examine some representative human face-to-face interaction data and find the recurrent patterns of groups. Moreover, we extend the force-directed motion (FDM) model with the forces derived from similarity distances within a hidden space to reproduce the recurrent group patterns and many key properties of face-to-face interaction networks. Furthermore, we demonstrate that the FDM model effectively predicts information-spreading behaviors under higher-order interactions. Finally, our results reveal that the recurrence of triangular groups inhibits the spread of information in face-to-face interaction networks, and the higher-order interactions will make this phenomenon more pronounced. These findings represent a significant advancement in the understanding of group formation and may open new avenues for research into the effects of group interactions on information propagation processes.

Figures

Figures reproduced from arXiv: 2502.09348 by the authors.

Figure 1
Figure 1. Recurrence of groups in real-world and simulated networks. (a-d) show the recurrent patterns of full triangles in different real-world networks. The gray dashed line separates the estimation and validation sections used in the model. (e-h) and (i-l) show the recurrent patterns of full triangles for the corresponding networks simulated by the FDM and AM models, respectively. In each figure, the purple lines represent… view at source ↗
Figure 2
Figure 2. Network properties of the real-world datasets and corresponding simulated networks. Each column corresponds to a real network. (a-d) show the distribution of contact duration between a pair of nodes. (e-h) show the distribution of interval time between consecutive contacts of edges. (i-l) show the distribution of time duration between full triangles. (m-p) show the distribution of interval time between consecutive f… view at source ↗
Figure 3
Figure 3. The time evolution of infected nodes’ densities in real and simulated networks. Each column represents a real network. The upper and lower rows correspond to the cases without higher-order interactions (i.e., β△ = 0.0) and with higher-order interactions (i.e., β△ = 0.3), respectively. Symbols and solid lines illustrate the outcomes of the simulations and the MMCA, respectively. All results were derived from 10,000 i… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The final spreading range ρ ∗ as a function of infection probability β. Each column stands for a real network. The first and second rows represent the results without and with higher-order interactions (i.e., β△ = 0.0 and β△ = 0.3), respectively. The third row shows th…
Figure 5
Figure 5. Figure 5: Recurrence of full triangles suppresses the information spreading. Each column stands for a real network. (a-d) show the null model’s recurrent patterns of full triangles. In each figure, the purple lines represent the recurrent full triangles, while the black ones cor…
Figure 6
Figure 6. Figure 6: Recurrence of full triangles inhibits the information spreading in the FDM and AM models. Each figure shows the ρ ∗ as a function of β for simulated networks by the FDM or AM model and their corresponding null models. The simulated networks are generated using the para…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 48 canonical work pages

  1. [1]

    Barrat, C

    A. Barrat, C. Cattuto, Face-to-Face Interactions, in: Social Phenomena: From Data Analysis to Models, Springer, 2015, pp. 37–57

    work page 2015

  2. [2]

    Masuda, R

    N. Masuda, R. Lambiotte, A Guide to Temporal Networks, World Scientific, 2016

    work page 2016

  3. [3]

    Holme, Modern temporal network theory: a colloquium, The European Physical Journal B 88 (9) (2015) 234

    P. Holme, Modern temporal network theory: a colloquium, The European Physical Journal B 88 (9) (2015) 234

    work page 2015

  4. [4]

    Holme, J

    P. Holme, J. Saramäki, Temporal networks, Physics Reports 519 (3) (2012) 97–125

    work page 2012

  5. [5]

    Barbosa, M

    H. Barbosa, M. Barthelemy, G. Ghoshal, C. R. James, M. Lenormand, T. Louail, R. Menezes, J. J. Ramasco, F. Si- mini, M. Tomasini, Human mobility: Models and applications, Physics Reports 734 (2018) 1–74

    work page 2018

  6. [6]

    Starnini, B

    M. Starnini, B. Lepri, A. Baronchelli, A. Barrat, C. Cat- tuto, R. Pastor-Satorras, Robust Modeling of Human Contact Networks Across Different Scales and Proximity-Sensing Tech- niques, in: Social Informatics, Springer, 2017, pp. 536–551

    work page 2017

  7. [7]

    Isella, J

    L. Isella, J. Stehlé, A. Barrat, C. Cattuto, J.-F. Pinton, W. Van den Broeck, What’s in a crowd? Analysis of face-to-face behav- ioral networks, Journal of Theoretical Biology 271 (1) (2011) 166–180

    work page 2011

  8. [8]

    Cattuto, W

    C. Cattuto, W. Van den Broeck, A. Barrat, V. Colizza, J.-F. Pinton, A. Vespignani, Dynamics of Person-to-Person Interac- tionsfromDistributedRFIDSensorNetworks, PLOSONE5(7) (2010) 1–9

    work page 2010

Show all 49 references
  1. [9]

    Starnini, A

    M. Starnini, A. Baronchelli, R. Pastor-Satorras, Modeling Hu- man Dynamics of Face-to-Face Interaction Networks, Physical Review Letters 110 (2013) 168701

    2013

  2. [10]

    Starnini, A

    M. Starnini, A. Baronchelli, R. Pastor-Satorras, Model repro- duces individual, group and collective dynamics of human con- tact networks, Social Networks 47 (2016) 130–137

    2016

  3. [11]

    M. A. R. Flores, F. Papadopoulos, Similarity Forces and Recur- rent Components in Human Face-to-Face Interaction Networks, Physical Review Letters 121 (2018) 258301

    2018

  4. [12]

    Iacopini, G

    I. Iacopini, G. Petri, A. Barrat, V. Latora, Simplicial models of social contagion, Nature Communications 10 (1) (2019) 2485

    2019

  5. [13]

    J. T. Matamalas, S. Gómez, A. Arenas, Abrupt phase transition of epidemic spreading in simplicial complexes, Physical Review Research 2 (2020) 012049. 9

    2020

  6. [14]

    X. Wang, J. Wu, Z. Yang, K. Xu, Z. Wang, M. Zheng, The correlation between independent edge and triangle degrees pro- mote the explosive information spreading, Physica A: Statistical Mechanics and its Applications 640 (2024) 129702

    2024

  7. [15]

    Battiston, E

    F. Battiston, E. Amico, A. Barrat, G. Bianconi, G. Ferraz de Arruda, B. Franceschiello, I. Iacopini, S. Kéfi, V. Latora, Y.Moreno, M.M.Murray, T.P.Peixoto, F.Vaccarino, G.Petri, The physics of higher-order interactions in complex systems, Nature Physics 17 (10) (2021) 1093–1098

    2021

  8. [16]

    Battiston, G

    F. Battiston, G. Cencetti, I. Iacopini, V. Latora, M. Lucas, A.Patania, J.-G.Young, G.Petri, Networksbeyondpairwisein- teractions: Structure and dynamics, Physics Reports 874 (2020) 1–92

    2020

  9. [17]

    Boccaletti, P

    S. Boccaletti, P. De Lellis, C. del Genio, K. Alfaro-Bittner, R. Criado, S. Jalan, M. Romance, The structure and dynam- ics of networks with higher order interactions, Physics Reports 1018 (2023) 1–64

    2023

  10. [18]

    Greene, D

    D. Greene, D. Doyle, P. Cunningham, Tracking the Evolution of Communities in Dynamic Social Networks, in: 2010 Interna- tional Conference on Advances in Social Networks Analysis and Mining, IEEE, 2010, pp. 176–183

    2010

  11. [19]

    G. F. de Arruda, G. Petri, Y. Moreno, Social contagion models on hypergraphs, Physical Review Research 2 (2020) 023032

    2020

  12. [20]

    M.T.Schaub, A.R.Benson, P.Horn, G.Lippner, A.Jadbabaie, Random Walks on Simplicial Complexes and the Normalized Hodge 1-Laplacian, SIAM Review 62 (2) (2020) 353–391

    2020

  13. [21]

    Carletti, F

    T. Carletti, F. Battiston, G. Cencetti, D. Fanelli, Random walks on hypergraphs, Physical Review E 101 (2020) 022308

    2020

  14. [22]

    P. S. Skardal, A. Arenas, Abrupt Desynchronization and Ex- tensive Multistability in Globally Coupled Oscillator Simplexes, Physical Review Letters 122 (2019) 248301

    2019

  15. [23]

    A. P. Millán, J. J. Torres, G. Bianconi, Explosive Higher-Order Kuramoto Dynamics on Simplicial Complexes, Physical Review Letters 124 (2020) 218301

    2020

  16. [24]

    Lucas, G

    M. Lucas, G. Cencetti, F. Battiston, Multiorder Laplacian for synchronization in higher-order networks, Physical Review Re- search 2 033410

  17. [25]

    Neuhäuser, A

    L. Neuhäuser, A. Mellor, R. Lambiotte, Multibody interactions and nonlinear consensus dynamics on networked systems, Phys- ical Review E 101 (2020) 032310

    2020

  18. [26]

    Sahasrabuddhe, L

    R. Sahasrabuddhe, L. Neuhäuser, R. Lambiotte, Modelling non- linear consensus dynamics on hypergraphs, Journal of Physics: Complexity 2 (2) (2021) 025006

    2021

  19. [27]

    Jusup, P

    M. Jusup, P. Holme, K. Kanazawa, M. Takayasu, I. Romić, Z. Wang, S. Geček, T. Lipić, B. Podobnik, L. Wang, W. Luo, T. Klanjšček, J. Fan, S. Boccaletti, M. Perc, Social physics, Physics Reports 948 (2022) 1–148

    2022

  20. [28]

    Alvarez-Rodriguez, F

    U. Alvarez-Rodriguez, F. Battiston, G. F. de Arruda, Y. Moreno, M. Perc, V. Latora, Evolutionary dynamics of higher-order interactions in social networks, Nature Human Be- haviour 5 (5) (2021) 586–595

    2021

  21. [29]

    P. Ji, J. Ye, Y. Mu, W. Lin, Y. Tian, C. Hens, M. Perc, Y. Tang, J. Sun, J. Kurths, Signal propagation in complex networks, Physics Reports 1017 (2023) 1–96

    2023

  22. [30]

    Kumar, S

    A. Kumar, S. Chowdhary, V. Capraro, M. c. v. Perc, Evolution of honesty in higher-order social networks, Physical Review E 104 (2021) 054308

    2021

  23. [31]

    Majhi, M

    S. Majhi, M. Perc, D. Ghosh, Dynamics on higher-order net- works: a review, Journal of The Royal Society Interface 19 (188) (2022) 20220043

    2022

  24. [32]

    R. I. M. Dunbar, The Anatomy of Friendship, Trends in Cog- nitive Sciences 22 (1) (2018) 32–51

    2018

  25. [33]

    Latora, V

    V. Latora, V. Nicosia, G. Russo, Complex Networks: Principles, Methods and Applications, Cambridge University Press, 2017

    2017

  26. [34]

    Vespignani, Twenty years of network science (2018)

    A. Vespignani, Twenty years of network science (2018)

    2018

  27. [35]

    Iacopini, M

    I. Iacopini, M. Karsai, A. Barrat, The temporal dynamics of group interactions in higher-order social networks, Nature Com- munications 15 (1) (2024) 7391

    2024

  28. [36]

    Cencetti, F

    G. Cencetti, F. Battiston, B. Lepri, M. Karsai, Temporal prop- erties of higher-order interactions in social networks, Scientific Reports 11 (1) (2021) 7028

    2021

  29. [37]

    X. Wang, L. Zhou, Z. Tian, A. Li, Evolutionary dynamics with temporal higher-order interactions, arXiv:2306.01300 (2024)

    2024

  30. [38]

    Moinet, R

    A. Moinet, R. Pastor-Satorras, A. Barrat, Effect of risk per- ception on epidemic spreading in temporal networks, Physical Review E 97 (2018) 012313

    2018

  31. [39]

    Gallo, C

    L. Gallo, C. Zappalà, F. Karimi, F. Battiston, Higher-order modeling of face-to-face interactions, arXiv:2406.05026 (2024)

    2024 arXiv

  32. [40]

    Sociopatterns, http://www.sociopatterns.org/

  33. [41]

    Vanhems, A

    P. Vanhems, A. Barrat, C. Cattuto, J.-F. Pinton, N. Khanafer, C. Régis, B.-a. Kim, B. Comte, N. Voirin, Estimating Potential Infection Transmission Routes in Hospital Wards Using Wear- able Proximity Sensors, PLOS ONE 8 (9) (2013) 1–9

    2013

  34. [42]

    Mastrandrea, J

    R. Mastrandrea, J. Fournet, A. Barrat, Contact Patterns in a High School: A Comparison between Data Collected Us- ing Wearable Sensors, Contact Diaries and Friendship Surveys, PLOS ONE 10 (9) (2015) 1–26

    2015

  35. [43]

    Stehlé, N

    J. Stehlé, N. Voirin, A. Barrat, C. Cattuto, L. Isella, J.-F. Pin- ton, M. Quaggiotto, W. Van den Broeck, C. Régis, B. Lina, P. Vanhems, High-Resolution Measurements of Face-to-Face Contact Patterns in a Primary School, PLOS ONE 6 (8) (2011) 1–13

    2011

  36. [44]

    M. J. Keeling, P. Rohani, Modeling Infectious Diseases in Hu- mans and Animals, Princeton University Press, 2008

    2008

  37. [45]

    Chowdhary, A

    S. Chowdhary, A. Kumar, G. Cencetti, I. Iacopini, F. Battiston, Simplicial contagion in temporal higher-order networks, Journal of Physics: Complexity 2 (3) (2021) 035019. 10 Supplementary Information for The recurrence of groups inhibits the information spreading under higher...

    2021 arXiv

  38. [46]

    Using the initial ranges determined above, we generate a set of parameter lists by a small increment

    Generating synthetic temporal networks. Using the initial ranges determined above, we generate a set of parameter lists by a small increment. Here we set the increment as ∆ L = 1, ∆ µ1 = 0.01, ∆F0 = 0.01 and ∆µ2 = 0.01. We would go through every value in the parameter lists an...

  39. [47]

    Next, we filter the parameters based on the average group count c

    Filtering Parameters: For each combination of parameters, we firstly keep it if the n and l values of the simulated networks fall within an error margin of 0 .2 with the corresponding values of the real network. Next, we filter the parameters based on the average group count c...

  40. [48]

    The main text shows our results with a combination of parameters

    Parameter Determination: Based on the filtering method described above, multiple parameter sets may meet the requirements for each data set. The main text shows our results with a combination of parameters. We also find that the results are robust if we choose another set of p...

  41. [49]

    M. A. R. Flores and F. Papadopoulos, Similarity Forces and Recurrent Components in Human Face-to-Face Interaction Networks, Physical Review Letters 121, 258301 (2018). 5 100 101 102 edge duration 10−6 10−4 10−2 100 distribution (e) Empirical FDM 101 102 edge interval 10−3 10−2...

    2018

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.