Dynamics of Collective Information Processing for Risk Encoding in Social Networks during Crises

Ali Mostafavi; Chao Fan; Fangsheng Wu

REVIEW 4 major objections 6 minor 13 references

Dynamics of Collective Information Processing for Risk Encoding in Social Networks during Crises

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Crisis social networks display stable collective information-processing laws.

desk verdict Useful descriptive baseline for crisis Twitter dynamics, but the causal claims about memory and tie strength outrun the analysis. read the letter →

arxiv 2412.17342 v1 pith:VCLUNIDP submitted 2024-12-23 cs.SI physics.data-an

Chao Fan , Fangsheng Wu , Ali Mostafavi This is my paper · ORCID

classification cs.SIphysics.data-an

keywords collectiveinformationprocessingriskcommunicationcrisissocialnetworkspower-lawdistributionPageRankinfluencetemporalinvariancespatialdecayTwitterdisasterdata

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

The paper sets out to show that collective information processing on social media during crises obeys stable statistical regularities even as the online network grows and the disaster evolves. Using about 13 million tweets from five different threats, it claims that user activity proportions, the mix of communication structures, and PageRank-based user influence all remain temporally invariant as the network grows each day. The influence scores converge to the same power-law distribution, with exponent $\alpha \approx 1.98$, in all five cases, and the paper argues these patterns do not depend on users' communication memory or tie strength because new edges dominate daily activity and edge weights are almost all 1. Spatially, communication frequency and medium response time decay with physical distance, so information exchange is concentrated near the threat area. If true, the stability would make crisis-time network forecasting and targeted intervention substantially easier.

What carries the argument

The central object is the cumulative daily human network built from Twitter retweet, reply, and quote interactions, with users as nodes and weighted directed activities as edges. Each day's network contains all nodes and edges from previous days, so temporal invariance is measured against a strictly growing graph. PageRank on this graph supplies the influence score, normalized by the second minimum bin and fitted with a maximum-likelihood power-law model $p(x) \sim C x^{-\alpha}$; the exponent $\alpha \approx 1.9825$ is the quantitative signature of stable influence. The spatial machinery is the top-100 location set with great-circle distances via the haversine formula, used to show that communication frequency and medium response time decay past the scale of a state. The causal work is inferential: because the daily share of new edges stays above 85% and edge weights cluster at 1, the invariance is attributed to the absence of memory and to weak ties.

What would settle it

Track each ordered pair of users across the days of the same five disasters and compare the activity mix, response times, and PageRank-influence distribution of pairs who interacted before with pairs who had no prior contact; if prior-contact pairs behave measurably differently, the paper's no-memory conclusion is contradicted.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the risk-encoding networks formed during crises are temporally invariant in their activity, structure, and influence, with user influence following a power-law distribution across all five disasters. Activity proportions for original posts, retweets, replies, and quotes become approximately normal after a Box-Cox transformation, and this normal shape holds as new users and edges join day by day. The shares of converging, reciprocal, and self-loop communication structures also stay stable, with more than 80% of users appearing only in converging structures as information sources. Normalized PageRank influence scores from the cumulative networks collapse onto a single power-law curve with exponent $\alpha \approx 1.9825$ by the final day of each disaster. The paper interprets these patterns as independent of communication memory and tie strength: newly created edges account for more than 85% of daily edges, and most edge weights equal 1. Spatially, communication is bounded by distance, with localized spikes in the same or nearby locations and weak transmission to distant places.

Load-bearing premise

The load-bearing assumption is that 'communication memory' can be measured by the daily share of newly created edges; if memory instead operates through repeated exchanges between the same users or continued sharing of the same information, the test would not detect it and the conclusion that memory does not affect the observed patterns would collapse.

Editorial extensions

If this is right

Emergency response agencies could identify the top roughly 2% of influential communicators early in a disaster, since the influence ranking stabilizes after the first day and persists as the network grows.
Misinformation mitigation can concentrate on these stable hubs, because most propagated information originates from them while ordinary users' original information accounts for only about 1% of communication flow.
Crisis communication models can treat activity and structure distributions as fixed, making day-by-day forecasts and simulations much simpler.
Platforms could deliberately add cross-region edges, for example recommending a popular risk message from one cluster to an influential user in another, to close the spatial transmission gaps the paper documents.

Reading between the lines

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

A natural extension the paper does not pursue is testing whether the same temporal and spatial regularities appear outside crises, such as in political or entertainment news; if they do, the mechanism would reflect a general property of collective attention rather than of threat encoding alone.
The memory conclusion could be stress-tested by simulating agent-based networks with explicit memory parameters and checking whether a memory-free model reproduces the observed $\alpha \approx 1.98$; this would give the paper's proxy-based argument a mechanistic footing.
The observed spatial boundary is likely co-produced by platform recommendation algorithms as well as by user preference; if so, the documented gaps could shrink when algorithms favor out-of-region content, which is a testable intervention.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Read this one if you want a quick sense of where crisis informatics is empirically. The genuinely useful part is the cross-case descriptive baseline: over five different threat types (airport outage, blizzard, wildfire, two hurricanes), cumulative Twitter networks show stable activity ratios, stable communication-structure ratios, and a stable heavy-tailed influence distribution once the networks reach a few days of growth. That is a real empirical pattern worth knowing, and the authors have the data (13M tweets, 5M users) and the patience to show it day by day. The spatial decay finding—same-city communication spikes and cross-region gaps—is also a reasonable descriptive contribution, even if it is presented without error bars or significance tests.

The soft spots are mostly at the level of interpretation. The stress-test is right about the memory claim: the 85% new-edges statistic cannot see pre-crisis interactions, so it cannot support the conclusion that "communication memory ... does not significantly affect" the patterns. That inference is a non-sequitur, and it is load-bearing for the abstract's universal framing. The normal-distribution claim for activity proportions is also qualified by the exclusion of the two extreme bins that contain over 90% of users; that is honest about what was done, but the abstract's "stable" wording overstates the coverage. The power-law fit (alpha ~ 1.98) is a fitted summary rather than a prediction; no alternative distributions are tested, and the lower bound is set at the second minimum bin, so the exponent is likely to be fragile. The tie-strength claim rests on a narrow edge-weight distribution, not on any direct measure of relational strength. These are fixable in a revision, but they currently make the mechanistic and universal claims in the title and abstract unsupported.

What the paper does well is set up a transparent pipeline: cumulative network construction, PageRank, Box-Cox transform, KDE, and MLE power-law fitting, with a limitations section that acknowledges platform, country, and feed-algorithm scope. The authors clearly know the surrounding literature. The overclaims are the problem, not the analysis style.

Verdict: this deserves a serious referee, but the referee should push hard on the memory/tie-strength section and demand robustness checks (alternative distributions, bootstrapped spatial decay, sensitivity to x_min and beta). It is not a desk reject; it is a revise-and-resubmit with the causal framing dialed back to descriptive.

Referee Report

4 major / 6 minor

Summary. This paper analyzes roughly 13 million tweets from five crisis events (Atlanta airport power outage, North American blizzard, Kincade wildfire, Hurricane Harvey, Hurricane Florence) to characterize collective information processing in social networks. Communication events are projected onto directed, weighted, cumulatively growing networks; the authors measure activity proportions, communication structures, PageRank-based influence, edge weights, response times, and spatial communication patterns. The central claims are that user activities and communication structures are temporally invariant as networks grow, that social influence follows a stable power-law distribution with exponent α≈1.9825, that these patterns are unaffected by people's memory and tie strength, and that communications are spatially localized with decay over distance. The paper concludes with intervention strategies for emergency communication and misinformation mitigation.

Significance. If the claims were fully supported, the paper would offer useful empirical regularities for modeling risk communication in disasters and for designing network interventions. The dataset is large and spans five distinct hazard types, which is a real strength for any cross-event generalization. The authors also provide a clear operationalization of several network metrics and report goodness-of-fit checks for the power-law claim. However, the manuscript's headline conclusions—especially the claims of invariance under memory and tie strength, and the universal normal/power-law descriptions—go beyond what the presented measurements actually establish. The core value of the paper is therefore as a descriptive, multi-case observational study; the causal and mechanistic statements need substantial revision and additional analysis to be justified.

major comments (4)

[Section 3, Fig. 2 and Section 2.2] The claim that activity proportions follow a normal distribution after Box-Cox transformation applies only after excluding the two extreme bins that 'account for more than 90% of the activities' (Section 3, first paragraph of Results). The abstract and conclusion state more generally that 'the activities and structures are stable' and that the normal distribution is a general pattern. As written, the analysis describes the minority of users with mixed activity types, not the full user population. The paper should either qualify the normality and invariance claims to explicitly exclude the dominant single-activity majority, or provide a complementary characterization of the full distribution that does not discard the majority of users.
[Section 3, memory paragraph (Fig. 5)] The abstract's claim that the identified patterns are 'not affected by people's memory' rests entirely on the observation that new edges constitute over 85% of edges created each day. This is a non-sequitur. The 'new edges' statistic counts edges absent from the crisis-period cumulative network; it cannot see interactions that occurred before the crisis, which is exactly the 'previous communications prior to the threat event' that the paper defines as memory. A pair of users who communicated frequently before the crisis and continue to interact during it will appear as a 'new' edge on the first crisis day. The high proportion of new edges is therefore compatible with strong pre-existing relational memory. The presented analysis cannot support the conclusion that memory does not affect activity or structure.
[Section 2.3 and Fig. 6] The power-law exponent α≈1.9825 is a fitted value, not a prediction from a generative model, and its reported stability is obtained using a lower bound chosen as 'the value of the second minimum bin' rather than a principled x_min selection such as the Clauset-Shalizi-Newman procedure. With this arbitrary cutoff, the claim that the influence distribution 'follows a power law' is a summary of a histogram after normalization, not a demonstrated universal law. The paper should either (a) derive the exponent from a stated mechanism (e.g., preferential attachment) and test the prediction, or (b) report the fit with a statistically justified x_min, the uncertainty in α, and a comparison against alternative heavy-tailed distributions. Without this, the 'stable power-law' claim is not as strong as the abstract presents it.
[Section 3, edge-weight analysis (Fig. 7)] The conclusion that tie strength does not affect the observed patterns is inferred from the fact that median edge weights are 1 and variances are less than 3. This describes the marginal distribution of edge weights; it does not test whether repeated interactions between the same users (or the presence of a few high-weight edges) alter activity dynamics, structure ratios, or PageRank values. Because PageRank is computed on unweighted edges, the analysis is structurally unable to detect an effect of tie strength on influence. The abstract's statement that the patterns are 'not affected by people's memory and ties' strength' is therefore not established. At minimum, the authors should compare weighted and unweighted versions of the network metrics or perform a regression-style test relating tie strength to activity and structure outcomes.

minor comments (6)

[Equation (1)] The word 'activties' in the definition of activity proportion is a typo; also the equation's denominator 'all activties of a user' should be 'all activities of a user'.
[Throughout, e.g., Figs. 9–10] The metric labeled 'medium response time' is aggregated by the median (the text says 'we use the medium response time'), so the label should be 'median response time' or the aggregation should be corrected.
[Section 3, fourth paragraph] The text states 'As shown in Fig. 7, we can find that while influential communicators only account for a very small percentage...' but Fig. 7 displays the distribution of edge weights, not the communication flows between influential and ordinary users. The correct reference appears to be Fig. 8. Please fix the cross-reference.
[Section 2.2] The Box-Cox transformation and the normal-distribution claim rely on a kernel bandwidth fixed at h=0.3 and a Gaussian kernel; the sensitivity of the normality conclusion to these choices is not reported. A brief robustness check (e.g., different bandwidths) would strengthen the claim.
[Section 2.4 and Figs. 9–10] The city and state distance thresholds (2 km and 100 km) are presented as 'the scale of a city' and 'the scale of a state' without a citation or justification; the conclusions about spatially bounded communication depend on these thresholds and should be justified or varied in a sensitivity analysis.
[Abstract and Conclusions] The phrase 'In addition, spatially localized communication spikes and global transmission gaps in the networks' is grammatically incomplete and should be rephrased.

Circularity Check

2 steps flagged · score 5.0 of 10

Tie-strength invariance is guaranteed by the unweighted PageRank construction, and memory invariance is built into the new-edge proxy; the temporal-stability core remains an independent empirical finding.

self definitional [Section 3, third results subsection (around Fig. 6 and Fig. 7), building on Eqs. (5)-(6)]
"Yet, the PageRank algorithm does not take the weight of the edges into account. This would raise an important question: would the weights of edges affect the influence of people in networks? We investigated the weights of edges and show the distributions for each threat case (Fig. 7 for Hurricane Harvey, and Supplementary Information Fig. S8 for the rest of the threats). We find that the medium weights of the edges for all threats are 1, and the variances are less than 3, while there are a few extreme weights that exist in the networks."

The PageRank influence scores are computed from the unweighted transition matrix M_ij in Eqs. (5)-(6), where the probability of moving along an edge is the ratio of that edge to the number of out-edges from the source node, not the edge weight. By construction, therefore, the influence scores cannot depend on edge weights. Concluding from the edge-weight distribution that 'the edge weights do not induce variation for the influence of people in the networks' is not an empirical finding; it restates a property already imposed by the algorithm. This step is load-bearing for the abstract's claim that the temporally invariant patterns are 'not affected by people's memory and ties' strength.'
self definitional [Section 3, second results subsection (paragraph after Fig. 4, introducing Fig. 5)]
"The communication memory of users in this context means that user's communication activities with other users is affected by their previous communications prior to the threat event. We examined the proportion of newly created edges to all edges created in each day during the threats. While the proportion of new edges decreases as the network grows, the new edges still account for more than 85% of the edges created each day (Fig. 5, and Supplementary Information Fig. 7)."

The stated quantity is memory of 'previous communications prior to the threat event,' but the measured quantity is the proportion of edges that are new within the crisis-period cumulative networks. A pair that communicated regularly before the crisis and resumes on the first crisis day contributes a 'new' edge on that day, so the >85% new-edge statistic is compatible with strong pre-existing relational memory. Since the proxy effectively defines memory as the absence of observed repeat edges inside the crisis window, the conclusion that memory does not significantly affect the patterns is an artifact of the operationalization rather than a test of the stated hypothesis. This step directly supports the abstract's claim that the patterns are 'not affected by people's memory.'

full rationale

The paper's central temporal-invariance results are empirical regularities: the Box-Cox transformed activity proportions, the communication-structure ratios, and the daily PageRank influence distributions are compared across five crisis datasets and reported as stable. That descriptive core is self-contained and is not circular merely because the power-law exponent alpha is obtained by maximum-likelihood fitting; the paper presents it as a fitted empirical characterization, not as a prediction from a stated generative model. No load-bearing uniqueness theorem or self-citation chain is invoked, so the self-citations in the references do not raise the score. The circularity is concentrated in two components of the abstract's blanket claim that the invariant patterns are 'not affected by people's memory and ties' strength.' First, the tie-strength conclusion is definitional: PageRank is computed on an unweighted transition matrix, so edge weights cannot, by construction, influence the resulting scores; the paper then interprets this built-in invariance as empirical evidence. Second, the memory conclusion is operationalized through the proportion of newly created edges, which cannot detect communication that occurred prior to the crisis even though the paper defines memory as 'previous communications prior to the threat event.' Both steps are load-bearing for the stated universality of the mechanisms, but they do not undermine the independent descriptive finding of temporal stability. Overall score 5 reflects partial circularity confined to the memory and tie-strength claims.

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

The central empirical regularities depend on several fitted parameters (α, λ, h, β, x_min, distance thresholds) and on domain assumptions equating Twitter actions with information processing and PageRank with influence. No new entities are introduced. The fitted parameters are not predictions from a model; they are summaries of the data.

free parameters (6)

power-law exponent α = ≈1.9825
Fitted via maximum likelihood (Clauset et al. 2009) to PageRank influence scores in daily cumulative networks; the claimed universal influence distribution depends on this fit.
Box-Cox λ = value maximizing log-likelihood
Chosen automatically to make transformed activity proportions approximately normal; the normality claim depends on this transformation.
KDE bandwidth h = 0.3
Chosen by hand for kernel density estimates; affects shape of reported distributions.
PageRank damping β = between 0.8 and 0.9
Set to a standard range; can change influence scores and the fitted power-law exponent.
lower bound of power-law fit (x_min) = second minimum bin of influence scores
Used to normalize influence scores and as the onset of power-law behavior; arbitrary choice affects the exponent and p-value.
city and state distance thresholds = approx. 2 km and 100 km
Used to describe spatial decay in communication frequency and response time; illustrative thresholds without statistical justification.

assumptions (4)

domain assumption Retweets, replies, and quotes represent social endorsement and information-processing behavior.
Stated in Section 2.3: 'User activities including retweet, reply, and quote are considered as a form of social endorsement or trust.'
domain assumption PageRank scores measure user influence in crisis communication networks.
The entire influence analysis (Section 2.3) treats PageRank as the quantitative measure of influence.
domain assumption Box-Cox transformed activity proportions are normally distributed for users with multiple activity types.
The paper asserts after transformation 'our data can be considered as normally distributed with constant variance and expectations' (Section 2.2) and then uses KDE and Q-Q plots without formal normality test.
standard math A single power-law model with p>0.01 from a KS test is sufficient to characterize the influence distribution.
The power-law testing follows Clauset et al. (2009) but the paper does not compare against alternative heavy-tailed distributions, despite that being part of the standard methodology.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Dynamics of Collective Information Processing for Risk Encoding in Social Networks during Crises." pith.science (2026). https://pith.science/paper/VCLUNIDP

@misc{pith2026241217342,
  author       = {Pith},
  title        = {Pith review of: Dynamics of Collective Information Processing for Risk Encoding in Social Networks during Crises},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VCLUNIDP}},
  note         = {Machine review of arXiv:2412.17342}
}

read the original abstract

Online social networks are increasingly being utilized for collective sense making and information processing in disasters. However, the underlying mechanisms that shape the dynamics of collective intelligence in online social networks during disasters is not fully understood. To bridge this gap, we examine the mechanisms of collective information processing in human networks during five threat cases including airport power outage, hurricanes, wildfire, and blizzard, considering the temporal and spatial dimensions. Using the 13MM Twitter data generated by 5MM online users during these threats, we examined human activities, communication structures and frequency, social influence, information flow, and medium response time in social networks. The results show that the activities and structures are stable in growing networks, which lead to a stable power-law distribution of the social influence in networks. These temporally invariant patterns are not affected by people's memory and ties' strength. In addition, spatially localized communication spikes and global transmission gaps in the networks. The findings could inform about network intervention strategies to enable a healthy and efficient online environment, with potential long-term impact on risk communication and emergency response.

Figures

Figures reproduced from arXiv: 2412.17342 by the authors.

**Figure 1.** The conceptual framework for characterizing collective information processing for encoding risks in social networks during crises. 2. Materials and Methods 2.1. Data collection and human network construction The Twitter data used in this study were collected via the Twitter Application Programming Interface (API) and all information analyzed was publicly available. The data include the text of the messages, posted t… view at source ↗

**Figure 2.** Probability density and Box-Cox transformation of user activity proportions for different activities: original post (A), retweet (B), quote (C), and reply (D), in the case of Hurricane Harvey. (see the distributions for other threat cases in Supplementary Materials Fig. S9-S12.) [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

**Figure 3.** Transformed probability density function of transformed activity proportions for different types of activities: (A) original post; (B) retweet; (C) quote; and (D) reply. Second, we measure the structures of the networks and quantify the effect of memory on network structures. The minimum units to characterize the communication patterns and information flow are the communication structures: converging, reciprocal, an… view at source ↗

Figures from the paper (5 more)

**Figure 4.** Figure 4: The proportions of communication structures in growing human networks for five disaster cases: (A) power outage at Atlanta airport, (B) blizzard in North American, (C) Kincade wildfire in California, (D) Hurricane Harvey in Houston, and (E) Hurricane Florence in North …

**Figure 6.** Figure 6: Complementary cumulative distribution function 𝑃𝑟(𝑥 ≥ 𝑋) of normalized influence scores of users for five disaster cases: (A) power outage at Atlanta airport, (B) blizzard in North American, (C) Kincade wildfire in California, (D) Hurricane Harvey in Houston, and (E) H…

**Figure 7.** Figure 7: The distribution of edge weights in each day during Hurricane Harvey. The area of each triangle represents the number of edges, and the height represents the range of possible values [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]

**Figure 9.** Figure 9: (A) The communication frequency distribution across top 100 locations identified from users’ profiles for Hurricane Harvey. (B) The relationship between communication frequency and location distances (in meters) during Hurricane Harvey. The first mark indicates the sca…

**Figure 10.** Figure 10: (A) The distribution of medium response time across top 100 locations identified from users’ profiles for Hurricane Harvey. (B) The relationship between medium response time and location distances (in meters) during Hurricane Harvey. The first mark indicates the scale…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 10 canonical work pages

[1]

Introduction Risks and threats are temporally evolving and spatially distributed (Banerjee et al., 2023; Liu & Fan, 2023; Xie et al., 2023). Rapidly and effectively processing situational information to inform decisions 2 and actions is essential for people under imminently risky and uncertain conditions (Fan & Mostafavi, 2019a; Kellner et al., 2023). For...

work page 2023
[2]

isolated

Materials and Methods 2.1. Data collection and human network construction The Twitter data used in this study were collected via the Twitter Application Programming Interface (API) and all information analyzed was publicly available. The data include the text of the messages, posted time stamps, user identifiers, location profiles, and retweet/reply/quote...

work page 1964
[3]

The occurrence of crisis events and their impacts trigger communication activities that grows the size of online social networks related to threat topics

Results First, we examined the activities in the growing networks. The occurrence of crisis events and their impacts trigger communication activities that grows the size of online social networks related to threat topics. Our analysis is conducted on a large corpus of risk messages on Twitter which allows people to post, retweet, reply, and quote messages...

work page 2023
[5]

(the proportions of new edges for other threats can be found in Supplementary Information Fig

Proportions of new edges among all edges in each day during Hurricane Harvey. (the proportions of new edges for other threats can be found in Supplementary Information Fig. S7) Third, we particularly look at how the social networks are formed and what are the main characteristics in the network, including user influence, tie strength and information flows...

work page 2009
[8]

Second, we measure the structures of the networks and quantify the effect of memory on network structures

Transformed probability density function of transformed activity proportions for different types of activities: (A) original post; (B) retweet; (C) quote; and (D) reply. Second, we measure the structures of the networks and quantify the effect of memory on network structures. The minimum units to characterize the communication patterns and information flo...

work page 2020
[10]

The values in the figure represent the percentages of the flow among all communication channels

Communication flows between influential users (top 2%) and ordinary users for five disaster cases: (A) power outage at Atlanta airport, (B) blizzard in North American, (C) Kincade wildfire in California, (D) Hurricane Harvey in Houston, and (E) Hurricane Florence in North Carolina. The values in the figure represent the percentages of the flow among all c...

work page 2021
[11]

Conclusion. This study provides quantitative empirical evidence needed to characterize the fundamental and universal mechanisms of collective information processing for risk encoding in social networks during crises. Specifically, the risk-encoded social networks are temporally invariant in terms of their network structures, user activities, and influence...

arXiv 2018
[29]

https://doi.org/10.1007/s41109-020-00271-5 Friedland, L., Joseph, K., Swire-Thompson, B., Grinberg, N., & Lazer, D. (2019). Fake news on Twitter during the 2016 U.S. presidential election. Science, 363(6425), 374–378. https://doi.org/10.1126/science.aau2706 GIS and Geospatial Professional Community. (2017). Distance on a sphere: The Haversine Formula. htt...

arXiv 2019

Show all 13 references

[2009]

In addition, to test the goodness-of-fit, we measured the corresponding p-values using the Kolmogorov-Smirnov test by generating 1,000 synthetic distributions (Noulas et al., 2012)

to fit the probability density function of user influence in daily human networks with a power-law distribution (𝑝(𝑥)~𝐶𝑥/) for the five selected threat cases. In addition, to test the goodness-of-fit, we measured the corresponding p-values using the Kolmogorov-Smirnov test by ...

2012
[2018]

Then, observational studies examined activity and structure patterns of information cascades that build the influence of information communicators

to identify influential communicators in human networks. Then, observational studies examined activity and structure patterns of information cascades that build the influence of information communicators. For example, Min et al.’s study indicates that the influence of the comm...

2015
[2020]

In fact, human networks are dynamic with newly joined people and edges during the threats (Comfort et al., 2004)

in uncovering the contribution of human influence to collective information processing, current research is limited to the patterns under stationary networks. In fact, human networks are dynamic with newly joined people and edges during the threats (Comfort et al., 2004). Howe...

2004
[2022]

and Nash equilibriums(Jackson, 2010), to approximate the real-world situations of information adoption. Moreover, recent studies (Fan & Mostafavi, 2019b) have realized the importance of the societal characteristics of people (e.g., social-economic contexts and interactions) an...

2010
[4787]

https://doi.org/10.1038/s41467-018-06930-7 Sosna, M. M. G., Twomey, C. R., Bak-Coleman, J., Poel, W., Daniels, B. C., Romanczuk, P., & Couzin, I. D. (2019). Individual and collective encoding of risk in animal groups. Proceedings of the National Academy of Sciences of the Unit...

2019

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Introduction Risks and threats are temporally evolving and spatially distributed (Banerjee et al., 2023; Liu & Fan, 2023; Xie et al., 2023). Rapidly and effectively processing situational information to inform decisions 2 and actions is essential for people under imminently risky and uncertain conditions (Fan & Mostafavi, 2019a; Kellner et al., 2023). For...

work page 2023

[2] [2]

isolated

Materials and Methods 2.1. Data collection and human network construction The Twitter data used in this study were collected via the Twitter Application Programming Interface (API) and all information analyzed was publicly available. The data include the text of the messages, posted time stamps, user identifiers, location profiles, and retweet/reply/quote...

work page 1964

[3] [3]

The occurrence of crisis events and their impacts trigger communication activities that grows the size of online social networks related to threat topics

Results First, we examined the activities in the growing networks. The occurrence of crisis events and their impacts trigger communication activities that grows the size of online social networks related to threat topics. Our analysis is conducted on a large corpus of risk messages on Twitter which allows people to post, retweet, reply, and quote messages...

work page 2023

[4] [5]

(the proportions of new edges for other threats can be found in Supplementary Information Fig

Proportions of new edges among all edges in each day during Hurricane Harvey. (the proportions of new edges for other threats can be found in Supplementary Information Fig. S7) Third, we particularly look at how the social networks are formed and what are the main characteristics in the network, including user influence, tie strength and information flows...

work page 2009

[5] [8]

Second, we measure the structures of the networks and quantify the effect of memory on network structures

Transformed probability density function of transformed activity proportions for different types of activities: (A) original post; (B) retweet; (C) quote; and (D) reply. Second, we measure the structures of the networks and quantify the effect of memory on network structures. The minimum units to characterize the communication patterns and information flo...

work page 2020

[6] [10]

The values in the figure represent the percentages of the flow among all communication channels

Communication flows between influential users (top 2%) and ordinary users for five disaster cases: (A) power outage at Atlanta airport, (B) blizzard in North American, (C) Kincade wildfire in California, (D) Hurricane Harvey in Houston, and (E) Hurricane Florence in North Carolina. The values in the figure represent the percentages of the flow among all c...

work page 2021

[7] [11]

Conclusion. This study provides quantitative empirical evidence needed to characterize the fundamental and universal mechanisms of collective information processing for risk encoding in social networks during crises. Specifically, the risk-encoded social networks are temporally invariant in terms of their network structures, user activities, and influence...

arXiv 2018

[8] [29]

https://doi.org/10.1007/s41109-020-00271-5 Friedland, L., Joseph, K., Swire-Thompson, B., Grinberg, N., & Lazer, D. (2019). Fake news on Twitter during the 2016 U.S. presidential election. Science, 363(6425), 374–378. https://doi.org/10.1126/science.aau2706 GIS and Geospatial Professional Community. (2017). Distance on a sphere: The Haversine Formula. htt...

arXiv 2019