Forecasting the Growth of Social Media Information Cascades: Towards Human-in-the-Loop Misinformation Triage

Ansh Gupta; Abhiram Gorle; Aayush Rajesh; Tsachy Weissman

cs.SIOct 5, 2026

Forecasting the Growth of Social Media Information Cascades: Towards Human-in-the-Loop Misinformation Triage

Authors: Ansh Gupta, Abhiram Gorle, Aayush Rajesh, Tsachy Weissman

Organizations: The Overlake School · Stanford University

Abstract

Limited review teams must identify which emerging claims are likely to keep growing before their eventual reach is known. We center early misinformation triage on this continuation-forecasting problem: predicting subsequent recorded propagation-tree growth from the first 30 minutes of activity. On FibVID, we compare early node count with structural depth entropy, temporal arrival entropy, and their pair while keeping all propagation trees from each original claim in one partition. Across 352 test trees from 59 claim groups separate from training, the combined model raises $R^2$ for log-transformed future growth from 0.307 to 0.323 and reduces log-MAE by 2.4% (95% claim-bootstrap CI, -0.4% to 5.2%). The gain is especially pronounced among 97 high-activity trees: $R^2$ rises from 0.248 to 0.395, Spearman's $ρ$ from 0.394 to 0.529, and log-MAE falls by 11.7% (95% CI, -1.9% to 23.9%). Complementing the 30-minute growth forecast, we analyze the first 15 replies in 563 PHEME threads. In this cohort, the 15th reply arrives after a median of 28.8 minutes; 52.0% reach the fixed reply prefix within 30 minutes and 71.6% within one hour. Even without the LLM-generated factual-accuracy dimension, the remaining stance, communicative, and affective state composition retains cross-event ranking signal (ROC-AUC 0.538); including that dimension increases ROC-AUC to 0.562. In a separate Check-COVID evaluation of 229 claims, reciprocal-rank fusion retrieves a gold evidence document within the top five for 74.2% of claims and within the top 20 for 94.3%; sentence reranking reaches Recall@20 of 58.1%. We propose an integrated human-review system that brings these early forecasts, response patterns, and retrieved evidence together for misinformation triage relying on the potential virality of claims.

Figures & tables

Panel of Figure 1 — Figure 1 : Proposed asynchronous early-triage workflow. The growth forecast is available at 30 minutes, while response context is added when 15 replies have arrived. The two predictive axes remain separate, and each contributes evidence to a human review queue.

Branch	Source	$n$	Unit	Evaluation role
Virality	FibVID	1,774	rooted propagation trees from 295 claims	1,422 training; 352 test trees, split by claim
Veracity	PHEME	563	binary-labeled threads with $\geq 15$ raw replies	event-held-out screen; descriptive sparse LDA
Retrieval	Check-COVID	229	official test claims	gold-evidence retrieval evaluation

Table 1 : Datasets used in the reported experiments. “Eligible” indicates that all task-specific observation and label requirements were met.

Panel of Figure 2 — Figure 2 : Illustration of structural depth and temporal bucketing in an early propagation tree.

Model	Log-MAE	$R^{2}$	Spearman $\rho$	Node MAE
Count only	1.2323	0.3067	0.5401	75.60
Count + structural entropy	1.2139	0.3196	0.5514	74.10
Count + temporal entropy	1.2334	0.3039	0.5401	75.61
Count + both entropies	1.2029	0.3227	0.5535	72.21

Table 2 : FibVID propagation-tree continuation ablation on all 352 test trees from 59 claim groups. Log-MAE and

R^{2}

use

\log(1+G_{c})

; node MAE uses the reconstructed number of subsequent recorded nodes. Errors are lower-is-better. Bold denotes the best value among the four configurations.

Model	Log-MAE	$R^{2}$	Spearman $\rho$	Node MAE
Count only	0.6400	0.2482	0.3941	76.17
Count + structural entropy	0.5882	0.3738	0.5239	72.94
Count + temporal entropy	0.6253	0.2396	0.4327	74.31
Count + both entropies	0.5651	0.3954	0.5288	67.22

Table 3 : FibVID propagation-tree continuation ablation on 97 high-activity test trees from 35 claims, defined by

n_{c}^{30}\geq 15

. Each model is fitted on the same complete training partition used in Table 2 . The subgroup is selected using only 30-minute activity.

Feature family	Candidates	ROC–AUC	PR–AUC	Bal. acc.	Macro-F1
Combined response, lexical, and structural	1,217	0.492	0.745	0.471	0.465
State composition	94	0.538	0.764	0.522	0.484
All response and lexical features	1,209	0.488	0.742	0.467	0.463
Prefix structure and timing	8	0.539	0.765	0.522	0.506
Entropy and transition dynamics	1,113	0.460	0.720	0.473	0.466

Table 4 : Primary PHEME event-grouped out-of-fold evaluation after excluding explicit factual-alignment-derived features. Each feature-family model is trained without the held-out events. Candidate counts are reported before within-fold variance filtering and sparse regularization. PR–AUC should be interpreted relative to the 73.5% true-class prevalence.

Figure 3 : Density of the 563 eligible PHEME threads projected onto the descriptive ten-feature sparse LDA axis. Positive values are oriented toward the true-associated distribution. Feature selection and fitting use this same complete sample, so the figure illustrates descriptive separation rather than held-out classification.

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

Event	False	True	Total
Sydney Siege	50	179	229
Ottawa Shooting	21	131	152
Charlie Hebdo	46	75	121
Germanwings crash	29	22	51
Ferguson	3	6	9
Gurlitt	0	1	1

Appendix

Table 6 : Event and veracity composition of the 563-thread PHEME sample.

Feature family	Candidates	ROC–AUC	PR–AUC	Bal. acc.	Macro-F1
State composition, without	94	0.538	0.764	0.522	0.484
State composition, with	104	0.562	0.774	0.539	0.501
All response and lexical, without	1,209	0.488	0.742	0.467	0.463
All response and lexical, with	1,305	0.543	0.765	0.545	0.532
Combined, without	1,217	0.492	0.745	0.471	0.465
Combined, with	1,313	0.546	0.767	0.549	0.536

Appendix

Table 7 : Event-held-out comparison without and with factual-alignment-derived features. Candidate counts are reported before within-fold variance filtering and sparse regularization.

Feature	Definition at the 30-minute observation cutoff
Observed node count	number of recorded nodes in the snapshot, including the root
Structural entropy	unnormalized Shannon entropy in bits over observed node depths, including the root; zero for a root-only snapshot
Temporal entropy	unnormalized Shannon entropy in bits over positive-time arrivals in six fixed five-minute bins; zero when no such arrival is observed

Appendix

Table 8 : The three predictors used in the FibVID propagation-tree continuation ablation. All are calculated at 30 minutes from the tree root.

Figure 4 : Predicted versus observed $\log(1+G_{c})$ for the 97 high-activity FibVID test propagation trees. Left: count-only baseline. Right: count plus structural and temporal entropy. Both models are fitted on 1,422 trees from 236 other claims. The dashed line denotes perfect prediction.

Figure 5 : Check-COVID evidence-document retrieval (left) and evidence-sentence reranking (right). The dashed line is the maximum sentence recall permitted by the 20-document candidate set.

Explore similar work

May 28, 2026cs.CL

CommunityFact: A Dynamic, Multilingual, Multi-domain Benchmark for Misinformation Detection in the Wild

Misinformation verification increasingly occurs in public, fast-moving, and multilingual online settings, where static benchmarks provide an incomplete measure of model reliability. We introduce CommunityFact, a refreshable benchmark for misinformation detection in the wild, with three major goals: coverage, granularity, and redistributability. This release contains 15,992 standalone claims across five languages and two domains. We evaluate ten LLMs under varying inference-time capabilities, including thinking and web-search. Our results show that closed-input verification remains challenging, web access yields the largest gains, and web-enabled LLMs' source-selection policies are systematically misaligned with the sources human Community Notes raters converge on -- a gap that closes through model-specific mechanisms of retrieval expansion or pruning. We further find substantial variation across language-domain slices and across the evidence ecosystems used by web-enabled systems. Beyond evaluation, CommunityFact positions Community Notes as a training signal for claim-conditioned source suggesters that could improve factual verification on novel claims.

Sahajpreet Singh, Insyirah Mujtahid, Min-Yen Kan +1

National University of Singapore

Misinformation Detection Large Language Model Reliability Multilingual Slices Recommendation Quality

Ask PDF

Aug 31, 2026cs.HC

One AI Signal, Many Human Judgments: A Bayesian Cascade Analysis of AI-based Credibility Indicators in Online Information Spread

Social media platforms increasingly use AI-based credibility indicators to help users judge misinformation. Unlike individual human-AI decision-making, these indicators are embedded in information spread: users see both an AI prediction and earlier judgments shaped by the same AI, and their own judgments may then enter the public history. Yet how to analytically characterize this process remains under-explored. We therefore introduce a social-learning lens for this setting by extending the classical Bayesian cascade model with the AI indicator as a shared public signal. The resulting Gateway condition compares the evidence from the AI prediction with users' private impressions. Through this view, we show that AI changes what public history means. Crowd agreement may reflect accumulated independent human evidence, or repeated dependence on the same AI prediction. This creates a preservation-correction trade-off: stronger reliance on AI can preserve correct predictions, but can also lock in incorrect ones by blocking corrective private impressions. We calibrate the model using human-subject data on news veracity judgments. Although the AI outperforms human users, the average user weights it below her own impression but above several peer judgments, while individual users vary from discounting the AI to relying on it enough to cascade. Simulations show that over-reliance on a weak AI is especially harmful, and that diversifying AI signals across users can better keep the crowd informative. We conclude with implications for understanding human-AI interaction in information spread and designing misinformation interventions.

Zhuoran Lu, Weilong Wang, Yangyang Yu +4

Purdue University West Lafayette, Indiana, USA · St John’s University New York, New York, USA · Stevens Institute of Technology Hoboken, New Jersey, USA +2

Credibility Misinformation Artificial Intelligence Systems Human Judgment

Ask PDF

Aug 31, 2026cs.LG

A Multi-Branch Feature Fusion Approach for Health Misinformation Detection and Propagation

This paper presents a multi-branch fusion framework for detecting and characterising the propagation of health misinformation in online social networks (OSNs). Grounded in the Elaboration Likelihood Model (ELM) and the Theory of Planned Behaviour (TPB), the model fuses transformer-based semantics with rhetorical cues, stance representations, and psychologically motivated proxies in a unified multi-task architecture. In addition to binary classification, we introduce the Cognitive Propagation Score (CPS), an interpretable post-hoc auxiliary score computed from psychologically motivated, text-derived cues capturing argument complexity, emotional intensity, and content-derived virality potential, to support diffusion-risk reasoning when engagement ground truth is incomplete or unavailable. Experiments on three benchmark datasets, Constraint, COVID--19_FNIR, and Monkeypox, show strong classification performance, achieving ROC--AUC up to 0.9999 on COVID--19_FNIR, while propagation-oriented ranking achieves near-perfect agreement when engagement-derived supervision is available (Monkeypox, Spearman's

ρ= 0.9952

) and similarly high ranking alignment under proxy-based supervision on COVID--19_FNIR (

ρ= 0.9954

). Compared with representative literature baselines, the fusion model improves detection on Constraint and COVID--19_FNIR, while Monkeypox remains more challenging, reflecting domain- and signal-specific differences. Ablation analysis further indicates that psychological and rhetorical branches provide complementary gains beyond semantic embeddings. Overall, the framework bridges cognitive theory and neural modelling to improve transparency and to support scalable misinformation monitoring, with future work required to validate CPS against human-centred diffusion judgements.

Mkululi Sikosana, Sean Maudsley-Barton, Oluwaseun Ajao

Department Of Computing And Mathematics, Manchester Metropolitan University, Manchester, UK

Misinformation Detection Misinformation Social Media False-Negative Rate Propagation Rhetoric+1

Ask PDF