Emotional Sequential Influence Modeling on False Information

Debashis Naskar; Sara Rodriguez Gonzalez; Subhashis Das

REVIEW 3 major objections 6 minor 11 references

Emotional Sequential Influence Modeling on False Information

T0 review · 3 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 E-USIM+ fuses GRU-encoded emotional history and neighbor emotions with BERT semantics to predict users' next emotions on rumor and non-rumor conversations, beating DeGroot, Voter, and E-USIM.

desk verdict A plausible incremental extension of the authors' own E-USIM, but the BERT term is never tied to a defined input window, so the headline gain may be content leakage rather than emotional influence. read the letter →

arxiv 2412.13668 v1 pith:2VLPRAB3 submitted 2024-12-18 cs.SI

Debashis Naskar , Subhashis Das , Sara Rodriguez Gonzalez This is my paper · ORCID

classification cs.SI

keywords emotionpredictionfalseinformationrumordetectionemotionalcontagionGRUBERTtemporalsocialnetworksequentialinfluence

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 proposes E-USIM+ (Emotional-based User Sequential Influence Model plus) and claims that a person's next emotion in a false-information conversation is predictably shaped by four sequential influences: what neighbors said emotionally before, what the person themselves felt before, what neighbors are saying now, and the person's accumulated emotional state. The model feeds these temporal signals through a gated recurrent unit (GRU), fuses the resulting memory vector with BERT-encoded semantic and social-context features, and classifies the next emotion into one of four regions: excitement, contentment, depression, or distress. On the 111-event Twitter rumor dataset (60 rumors, 51 non-rumors), E-USIM+ reaches 0.6728 accuracy on rumor conversations and 0.7187 on non-rumor conversations, outperforming DeGroot, Voter, and the earlier E-USIM. If this is right, emotional contagion in online rumor threads can be tracked at the individual level from message timing and content alone.

What carries the argument

The load-bearing object is the temporal tweet chain: for each user, outgoing posts are interleaved with incoming posts from neighbors, and each post carries one of four emotion regions from the circumplex model of affect. From these chains the model derives four influence signals—prior neighboring, prior personal, current neighboring, and historical—and feeds them into a GRU whose hidden state $C_{u,i}(t)$ acts as a running emotional memory. The GRU's reset and update gates decide what to keep and what to drop from the emotional past. The semantic branch uses BERT's multi-head self-attention to encode tweet words plus social structural features into $F_u(i)$, and the final vector $H_{u,i}(t) = (C_{u,i}(t) \oplus F_u(i))$ is passed to a softmax layer for classification.

What would settle it

Randomly permute the order of incoming neighbor tweets within each user's inter-post gap and retrain E-USIM+; if accuracy stays roughly the same, the temporal ordering the model exploits is not carrying the prediction, and the claimed sequential influence is an artifact of content features.

Watch

Extended reading notes

Core claim

The central claim is that historical emotional state is the missing ingredient in sequential influence models of emotion on false information. E-USIM+ replaces the static historical-influence term of E-USIM with the GRU's internal state $C_{u,i}(t)$, updated from prior neighboring influence $PN_{u,i}(t)$, prior personal influence $PP_{u,i}(t)$, and current neighboring influence $CN_{u,i}(t)$. That temporal state is concatenated with a BERT semantic representation $F_u(i)$ of tweet content and social structure, and a softmax layer outputs probabilities over the four emotion regions. The reported result is that E-USIM+ beats E-USIM by about five accuracy points on both rumor and non-rumor threads, with the largest F-measure gains on distress and excitement.

Load-bearing premise

The model assumes that emotions in tweets a user receives from neighbors between their own posts, together with the user's own emotional history, are what determine the user's next emotion; it does not test whether outside events, topic, or network structure are the real drivers.

Editorial extensions

If this is right

Models that ignore a user's own emotional history, such as DeGroot and Voter, are leaving usable signal on the table for rumor emotion prediction.
The four-region emotion mapping is sufficient to capture differences in how rumors and non-rumors propagate emotionally, since the same architecture yields different accuracy and F-measure profiles on the two conversation types.
Conversation tempo matters: the construction of incoming-neighbor windows between a user's outgoing posts is what creates the sequential signal, so the model is sensitive to the timing of messages.
Combining lexical emotion coverage with BERT semantic features lets the model assign emotions to tweets that contain no explicit emotion word.

Reading between the lines

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

A testable extension the paper does not run is a shuffled-timestamps control: randomizing the order of incoming neighbor tweets within each gap while keeping content identical would reveal whether the temporal influence signal is causal or just correlated with content.
The same GRU memory state could be used as a feature for predicting retweet or sharing decisions, since emotional contagion is believed to drive sharing behavior; this would connect emotion prediction directly to spread prediction.
Because the model relies only on message timing and emotion labels, it could be ported to other languages or platforms by re-labeling emotions with the same four quadrants, though the BERT branch would need retraining on new text.
The paper's implicit claim that influence is mostly local—neighbors and self—implies that platform feed design, which controls who appears in a user's stream, could directly change emotional dynamics in rumor threads.

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

Short version: this is a modest extension of the authors' earlier E-USIM model, and the stress-test concern is real. The paper never says what the BERT semantic vector Fu(i) sees. If it encodes the tweet whose emotion is the prediction target, then Table 1 mostly shows content classification and the GRU influence chain is bypassed.

What is new: E-USIM+ adds BERT semantic features to the GRU-based temporal emotion model and reports a fresh evaluation on the Kwon rumor dataset. The four-quadrant Russell emotion space and the temporal chain construction come from E-USIM, so the delta is the fusion of content/social features with the emotion-sequence GRU. The method is described well enough for someone with the existing pipeline to reimplement, though no code or hyperparameters are given.

What it does well: the paper is honest about the incremental nature, and the GRU equations are standard. The comparison is at least directional on the data: E-USIM+ beats DeGroot, Voter, and the re-implemented E-USIM on both rumor and non-rumor accuracy.

Soft spots: Table 1 has no confidence intervals, significance tests, or repeated runs, and it rests on one filtered Twitter dataset. DeGroot and Voter are weak baselines, and the E-USIM baseline is described as a 'simplified version' of the authors' own model, so the comparison is not strong. The bigger problem is Section 2.3/2.4: Fu(i) shares the index i with the target Su(i), and the paper never states whether it includes the content of the target tweet. Without that definition, and without a content-only ablation, the improvement cannot be attributed to historical temporal influence. This is a missing input-output definition, not just a missing robustness check. The writing also has typesetting errors that make the temporal chain hard to follow. The citation pattern is not what bothers me: leaning on [6] is normal for an incremental extension. The issue is that the baseline is called simplified.

Who this is for: researchers working on emotion forecasting in rumor threads, who might use this as a baseline method. I would not cite it as evidence until the leakage question is resolved.

Recommendation: I would desk reject the current form. The missing definition, no code, no statistics, and one dataset are too weak to justify referee time. If the authors resubmit with an explicit statement that Fu(i) is built only from prior/neighbor content, a content-only ablation, and significance tests, it becomes a reasonable short-paper submission.

Referee Report

3 major / 6 minor

Summary. The paper proposes E-USIM+, an extension of the authors' earlier E-USIM model, to predict a user's future emotional state on rumour and non-rumour tweets in online social networks. The model combines a GRU over temporal sequences of emotional influence signals (prior neighbouring, prior personal, current neighbouring, historical) with BERT-derived semantic and social structural features, fused as H_{u,i}(t) = C_{u,i}(t) ⊕ F_u(i), followed by a softmax classifier over four emotional regions. Experiments on the Kwon et al. (2017) dataset report accuracy improvements over DeGroot, Voter, and E-USIM baselines for both rumour and non-rumour conversations (e.g., rumour accuracy 0.6728 vs. 0.6234 for E-USIM).

Significance. If the reported gains are real and the model is specified without leakage, the paper would make a useful contribution to emotion dynamics and misinformation research by showing that combining temporal emotional influence with semantic content features improves future emotion prediction. The paper uses a publicly available dataset, compares against three baselines, and extends the authors' prior model, which provides continuity. However, the central empirical claim rests on a single dataset, a single run, and a model specification that is ambiguous about the alignment between the BERT-encoded content and the prediction target. The absence of code, ablations, and statistical tests limits the strength of the evidence, and the paper currently does not establish that the improvement is attributable to temporal emotional influence rather than content leakage or an uneven comparison.

major comments (3)

[§2.3–2.4, Eq. (5)–(6)] The most load-bearing issue is that the paper never states whether F_u(i) is the BERT embedding of the user's own tweet i whose emotion S_u(i) is the prediction target, or an embedding of preceding neighbour tweets and the user's own prior tweets. In §2.4 the model predicts P(S_u(i)|H_{u,i}(t)) with H_{u,i}(t) = C_{u,i}(t) ⊕ F_u(i); §2.3 defines F_u(i) as 'the learned tweet semantic representation vector' from 'the tweet content within the length n'. If F_u(i) encodes the very tweet whose emotion is the label, then the label can be inferred directly from content and the GRU influence chain is bypassed, making the comparison against E-USIM and the content-free DeGroot/Voter baselines uninformative. The paper must specify the exact alignment (which tweets and which timestamps are encoded in F_u(i)) and must report a content-only baseline (BERT features without the temporal GRU) and an ablation without content features to disentangle the contribution of temporal influence from content semantics.
[§3, Table 1] The reported accuracy differences, e.g., rumour accuracy 0.6728 vs. 0.6234 for E-USIM, are presented without any uncertainty quantification. There are no confidence intervals, no significance tests, no standard deviations over repeated runs, and only one dataset. Since the improvements are modest and the dataset is filtered to users with at least two messages and tweets that show emotion, the generalizability of the claim 'integrating historical information into an influential model is better' is not established. The authors should provide error bars, significance tests, and ideally results on additional datasets or at least a discussion of the filtering's impact on scope.
[§2.2, Eq. (4)] The temporal chain construction assumes that incoming neighbour tweets received between a user's posts, together with the user's own historical emotional sequence, are sufficient to determine the user's next emotion. This is a strong causal assumption. The paper does not control for exogenous events, topic, network structure, or text content beyond the BERT features, and does not rule out the possibility that temporal co-occurrence reflects common response to an external event rather than social influence. The authors should at least discuss this limitation and, ideally, add a control experiment (e.g., permuting the order of incoming tweets or using a shuffled-timestamp baseline) to test whether the temporal ordering, rather than mere co-occurrence, is responsible for the predictive gains.

minor comments (6)

[Abstract] The abstract contains a grammatical error: 'we systematically examining' should be 'we systematically examine'.
[§2, Fig. 1] Figure 1 is essentially unreadable in the provided manuscript; the labels and the flow are illegible. A larger, clearly annotated figure is needed.
[§2.1] The notation for the temporal tweet chain (e.g., arrows for incoming/outgoing tweets, the subscripts and superscripts) is difficult to parse. The authors should provide a clean example with concrete timestamps and users to show how the chain is built.
[§2.2] The definitions of the gates r_{u,i}(t) and g_{u,i}(t) in Eqs. (1)–(2) use nonstandard symbols and the text does not define the reset and update gates explicitly; a standard GRU formulation with clear gate names would improve readability.
[§3] No details are provided on hyperparameters, number of GRU layers, hidden dimensions, BERT version, or training configuration, making the experiments hard to reproduce. The authors should include a reproducibility statement or a hyperparameter table.
[References] Reference [6] is the authors' own prior E-USIM paper; the relation between E-USIM and E-USIM+ should be stated more explicitly in the text, not only in the comparison table.

Circularity Check

1 steps flagged · score 6.0 of 10

E-USIM+ feeds the BERT-encoded content of the very tweet whose emotion is the label into Hu,i(t), so the temporal-influence claim is partly by construction; score 6.

self definitional [Sections 2.1, 2.3, 2.4 (Eqs. (5) and (6)) and the fusion definition Hu,i(t) = (Cu,i(t) ⊕ Fu(i))]
"The each user u ∈ V , his/her posting sequence contains PS u = < Mu(i), Su(i), tu(i) > where a tweet message Mu(i) with emotion state Su(i) at time tu(i) posted by u. ... The tweet content within the length n is represented as a set of words Q={w1, w2, ..., wn} ... the learned tweet semantic representation vector Fu is obtained as follows: Fu(i) = max(0, TWa + ba) (5) ... These two feature vectors are fused to obtain a final feature vector set Hu,i(t) = (Cu,i(t) ⊕ Fu(i)). ... P (Su(i)|Hu,i(t)) = Softmax(V Hu,i(t) + b) (6)"

Eq. (6) predicts Su(i) given Hu,i(t), and Hu,i(t) is the concatenation of the GRU state Cu,i(t) with Fu(i). Fu(i) is defined in Eq. (5) as the learned semantic representation of 'the tweet content within the length n' for the same indexed tweet i, while Su(i) is defined in Sec. 2.1 as the emotion state of that same tweet Mu(i). Thus the content of the tweet whose emotion is the prediction target is already an input feature through BERT, so the softmax can assign the emotion from Fu(i) directly, bypassing the GRU influence chain in Eqs. (1)-(4).

full rationale

The central derivation is not circular in the usual fitting sense: the emotion labels are observed, the model is trained on them, and evaluation is on a held-out temporal split. However, the paper's own equations make the prediction partly self-definitional: the target label Su(i) is the emotion state of tweet Mu(i), and the feature vector Hu,i(t) used to predict Su(i) contains Fu(i), the BERT/Transformer semantic encoding of 'the tweet content within the length n' for that same tweet index. Unless Fu(i) is explicitly restricted to past or neighbouring tweets, the model can read the answer from the content feature, so the GRU-based historical influence mechanism is not independently validated. The use of the authors' earlier E-USIM model [6] as a baseline is not by itself circular, but it does mean the conclusion 'integrating historical information into an influential model is better' rests on a comparison with an unpublished-in-this-paper variant and on a model that may be leaking the target tweet's content. A content-only BERT ablation or a clearly stated exclusion of the target tweet's content from Fu(i) would resolve the ambiguity; without that, the claimed contribution is partially forced by the input definition.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The paper contributes a supervised architecture, not a derivation. Its central claim rests on domain assumptions about emotion labeling, temporal influence, network edges, and the adequacy of the reported evaluation. The model's many trainable parameters are fitted, but the paper reports no hyperparameters, so the fit cannot be inspected.

free parameters (2)

GRU/BERT architecture and training hyperparameters = not reported (GRU hidden dimension c, BERT variant, learning rate, batch size, epochs)
The reported accuracy depends on these choices; without them the comparison cannot be reproduced or assessed for overfitting.
Emotion region granularity = 4 quadrants (from Russell's 16 states)
Section 2.2 reduces the 16 circumplex states to four quadrants; this choice defines the classification targets and is not varied.

assumptions (5)

domain assumption Tweet emotions can be mapped to Russell's four quadrants (excitement, contentment, depression, distress) using EmoLex and the circumplex model.
Section 2.2 defines emotion states and the four quadrants; the accuracy of the entire label set depends on this mapping.
domain assumption Incoming neighbor tweets between a user's outgoing posts causally shape the user's next emotion.
Section 2.2 builds Xu,i(t) from prior neighboring, prior personal, and current neighboring emotion signals and updates the GRU state in Eq. 4; no causal controls are provided.
domain assumption The social network follow graph G=(V,E) is the channel through which emotional influence propagates.
Section 2.1 defines the neighbor set Zu from edges; the model assumes influence flows along following relationships.
domain assumption Pretrained BERT yields a semantic vector Fu(i) that captures content and social-context features needed for emotion prediction.
Section 2.3 uses BERT's multi-head self-attention output as the semantic feature vector; no BERT version or fine-tuning details are given.
standard math The GRU equations (1)-(4) are a valid model of sequential emotion state updates.
Section 2.2 applies the standard GRU recurrence; this is an accepted architectural primitive, so it is a background assumption rather than an ad hoc invention.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Emotional Sequential Influence Modeling on False Information." pith.science (2026). https://pith.science/paper/2VLPRAB3

@misc{pith2026241213668,
  author       = {Pith},
  title        = {Pith review of: Emotional Sequential Influence Modeling on False Information},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2VLPRAB3}},
  note         = {Machine review of arXiv:2412.13668}
}

read the original abstract

The extensive dissemination of false information in social networks affects netizens social lives, morals, and behaviours. When a neighbour expresses strong emotions (e.g., fear, anger, excitement) based on a false statement, these emotions can be transmitted to others, especially through interactions on social media. Therefore, exploring the mechanism that explains how an individuals emotions change under the influence of a neighbours false statement is a practically important task. In this work, we systematically examining the publics personal, interpersonal, and historical emotional influence based on social context, content, and emotional based features. The contribution of this paper is to build an emotionally infused model called the Emotional based User Sequential Influence Model(EUSIM) to understand users temporal emotional propagation patterns and predict future emotions against false information.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

11 extracted references · 11 canonical work pages

[6]

In: Proceedings of the 18th Inter - national Conference on Advances in Mobile Computing & Multimedia

Naskar, D., Onaindia, E., Rebollo, M., Singh, S.R.: Predicting emotion dynamics sequence on twitter via deep learning approach. In: Proceedings of the 18th Inter - national Conference on Advances in Mobile Computing & Multimedia. pp. 20 –24 (2020)

work page 2020
[1]

In: Proceedings of the 2017 ACM on Conference on Information and Knowledge Management

Chen, C., Wang, Z., Li, W.: Modeling opinion influence with user dual identity. In: Proceedings of the 2017 ACM on Conference on Information and Knowledge Management. pp. 2019–2022. Singapore (2017)

work page 2017
[2]

In: Proceedings of the 15th ACM Web Science Conference 2023

Hosseini, A.S., Staab, S.: Emotional framing in the spreading of false and true claims. In: Proceedings of the 15th ACM Web Science Conference 2023. pp. 96–

work page 2023
[3]

PloS one 12(1), e0168344 (2017)

Kwon, S., Cha, M., Jung, K.: Rumor detection over varying time windows. PloS one 12(1), e0168344 (2017)

work page 2017
[4]

Information Fusion p

Liu, Z., Zhang, T., Yang, K., Thompson, P., Yu, Z., Ananiadou, S.: Emotion de - tection for misinformation: A review. Information Fusion p. 102300 (2024)

work page 2024
[5]

In: Proceedings of the NAACL HLT 2010 Workshop on Computational Approaches to Analysis and Generation of Emotion in Text

Mohammad, S., Turney, P.: Emotions evoked by common words and phrases: Using Mechanical Turk to create an emotion lexicon. In: Proceedings of the NAACL HLT 2010 Workshop on Computational Approaches to Analysis and Generation of Emotion in Text. pp. 26–34. ACL, Los Angeles, CA (2010)

work page 2010
[7]

EPJ Data Science 10(1), 51 (2021)

Pröllochs, N., Bär, D., Feuerriegel, S.: Emotions in online rumor diffusion. EPJ Data Science 10(1), 51 (2021)

work page 2021
[8]

Journal of personality and social psy - chology 39(6), 1161 (1980)

Russell, J.A.: A circumplex model of affect. Journal of personality and social psy - chology 39(6), 1161 (1980)

work page 1980

Show all 11 references

[9]

In: 2020 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM)

Zaeem, R.N., Li, C., Barber, K.S.: On sentiment of online fake news. In: 2020 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM). pp. 760–767. Online (2020)

2020
[10]

this is fake! shared it by mistake

Zhou, X., Shu, K., Phoha, V.V., Liu, H., Zafarani, R.: “this is fake! shared it by mistake”: Assessing the intent of fake news spreaders. In: Proceedings of the ACM Web Conference 2022. pp. 3685–3694. Online (2022)

2022
[106]

Austin, TX, USA (2023)

2023

Pith tools

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

[1] [6]

In: Proceedings of the 18th Inter - national Conference on Advances in Mobile Computing & Multimedia

Naskar, D., Onaindia, E., Rebollo, M., Singh, S.R.: Predicting emotion dynamics sequence on twitter via deep learning approach. In: Proceedings of the 18th Inter - national Conference on Advances in Mobile Computing & Multimedia. pp. 20 –24 (2020)

work page 2020

[2] [1]

In: Proceedings of the 2017 ACM on Conference on Information and Knowledge Management

Chen, C., Wang, Z., Li, W.: Modeling opinion influence with user dual identity. In: Proceedings of the 2017 ACM on Conference on Information and Knowledge Management. pp. 2019–2022. Singapore (2017)

work page 2017

[3] [2]

In: Proceedings of the 15th ACM Web Science Conference 2023

Hosseini, A.S., Staab, S.: Emotional framing in the spreading of false and true claims. In: Proceedings of the 15th ACM Web Science Conference 2023. pp. 96–

work page 2023

[4] [3]

PloS one 12(1), e0168344 (2017)

Kwon, S., Cha, M., Jung, K.: Rumor detection over varying time windows. PloS one 12(1), e0168344 (2017)

work page 2017

[5] [4]

Information Fusion p

Liu, Z., Zhang, T., Yang, K., Thompson, P., Yu, Z., Ananiadou, S.: Emotion de - tection for misinformation: A review. Information Fusion p. 102300 (2024)

work page 2024

[6] [5]

In: Proceedings of the NAACL HLT 2010 Workshop on Computational Approaches to Analysis and Generation of Emotion in Text

Mohammad, S., Turney, P.: Emotions evoked by common words and phrases: Using Mechanical Turk to create an emotion lexicon. In: Proceedings of the NAACL HLT 2010 Workshop on Computational Approaches to Analysis and Generation of Emotion in Text. pp. 26–34. ACL, Los Angeles, CA (2010)

work page 2010

[7] [7]

EPJ Data Science 10(1), 51 (2021)

Pröllochs, N., Bär, D., Feuerriegel, S.: Emotions in online rumor diffusion. EPJ Data Science 10(1), 51 (2021)

work page 2021

[8] [8]

Journal of personality and social psy - chology 39(6), 1161 (1980)

Russell, J.A.: A circumplex model of affect. Journal of personality and social psy - chology 39(6), 1161 (1980)

work page 1980