Pith. sign in

REVIEW 3 major objections 6 minor 27 references

A Dashboard Approach to Monitoring Mpox-Related Discourse and Misinformation on Social Media

T0 review · 3 major / 6 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 This paper claims that an interactive dashboard can surface near-real-time shifts in mpox discourse, and that in its 2023–2024 data tweet volume rose sharply after the CDC's August 2024 designation while cynicism became the dominant…

desk verdict A well-intentioned dashboard paper whose central sentiment claims rest on an unnamed, unvalidated labeling step — useful as a design report, not yet as an empirical study. read the letter →

arxiv 2505.20584 v1 pith:C5R5PHJH submitted 2025-05-26 cs.SI cs.CYcs.HC

Linfeng (Leon) Zhao , Rishul Bhuvanagiri , Blake Gonzales , Kellen Sharp , Dhiraj Murthy This is my paper · ORCID
classification cs.SIcs.CYcs.HC
keywords mpoxdashboardmisinformationsocialmediahealthcommunicationsentimentanalysistopicclusteringpublicsurveillance
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 argues that an interactive social-media dashboard can support local public health agencies by making mpox-related discourse and misinformation visible in near real time. It reports that after the CDC designated mpox an emerging threat in August 2024, dashboard-recorded tweet volume rose markedly compared with 2023. It further reports that in the more recent data, cynicism toward public health institutions became the dominant sentiment in mpox discussions on X (formerly Twitter). If these observations hold, public health communicators would have an early signal of both rising attention and falling trust, allowing them to intervene before misinformation solidifies. The paper is a proof-of-concept application rather than a validation study.

What carries the argument

The carrying mechanism is the dashboard itself: a Streamlit web application that ingests filtered mpox tweets from three sources, presents searchable tables, lets users filter by keywords and engagement metrics such as like, reply, and retweet counts, and plots a time series of topic and sentiment clusters. The clustering relies on an algorithm that labels each tweet into thematic categories including cynicism, COVID-19 comparisons, government action, and misinformation; the proportion of tweets in each category per day is then shown as a color-coded scatterplot.

What would settle it

Take a random sample of the dashboard's 2024 mpox tweets, have independent human annotators label each tweet's sentiment and topic without seeing the dashboard's cluster labels, and compare the two distributions; if the human-labeled data does not show cynicism as the dominant category, or shows the algorithm systematically mislabels tweets, the central claim fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that a researcher-focused dashboard built from 2023–2024 mpox-related tweets can reveal real-time changes in public discourse, and that when applied to its collected data it recorded a marked increase in tweet volume after the CDC designated mpox an emerging virus in August 2024. The same dashboard analysis found that cynicism, defined as distrust in public health institutions and traditional news sources, became the dominant sentiment in the more recent mpox-related discussions, a pattern the authors interpret as a warning that audiences may be increasingly receptive to unofficial or misleading health information.

Load-bearing premise

The central finding that cynicism dominates recent mpox discussion rests entirely on the unnamed topic-and-sentiment labeling algorithm, whose accuracy is never tested, so if that algorithm mislabels tweets the main result collapses.

Editorial extensions

If this is right

  • Local public health agencies could use the dashboard to detect when discourse volume jumps after official announcements, enabling faster communication responses.
  • If cynicism is truly dominant, health communication strategies may need to prioritize trust repair over simply providing more facts.
  • The keyword and engagement filters allow researchers to track specific misinformation narratives and see which ones gain traction.
  • The cluster graph could reveal when COVID-19 comparisons or government-action narratives spike, helping agencies tailor messages to current concerns.

Reading between the lines

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

  • If the volume surge is real, the same dashboard pattern could serve as a generic early-warning signal for other emerging pathogens, though the paper does not control for platform-wide volume changes or differences in data collection between years.
  • The dominance of cynicism may be specific to X's user base or driven by particular news cycles; applying the same pipeline to Reddit or Facebook data would test whether the mood shift reflects broader public sentiment.
  • A natural extension would be to use the dashboard's keyword filters in an intervention study: if agencies respond to cynicism-heavy clusters with trust-building messages, one could measure whether the proportion of cynicism-labeled tweets falls in subsequent weeks.
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.

Referee Report

3 major / 6 minor

Summary. The paper describes a Streamlit-based dashboard for searching and visualizing mpox-related tweets, intended for local public health agencies. It combines three datasets (a UT Austin Computational Media Lab Spritzer-stream sample, a publicly available May 2022 Monkeypox dataset, and 3,000 tweets collected via Zeeschuimer in 2024), applies a Python filtering pipeline, and visualizes keyword trends, engagement metrics, and a color-coded cluster graph of sentiment/topic categories over time. The paper's central empirical claims are that the dashboard recorded a marked increase in tweet volume in 2024 compared with 2023, and that cynicism became the dominant sentiment in mpox-related discussions. The manuscript is primarily a system/tool description with an accompanying literature review, and it does not report quantitative validation of its measurement pipeline.

Significance. If the central claims were quantitatively supported, the dashboard would be a useful proof-of-concept for real-time infodemic monitoring by public health stakeholders, complementing prior tools such as PoxVerifi and the Johns Hopkins COVID-19 dashboard. The paper also usefully situates the work within the misinformation and health-communication literature. However, as written, the two headline findings—a marked 2024 versus 2023 volume increase and the dominance of cynicism—are not backed by reported counts, error bars, or any validation of the sentiment/topic labeling algorithm. The contribution is therefore better described as a dashboard demonstration than as a validated empirical study of mpox discourse.

major comments (3)
  1. [§5.2] The sentiment/topic labeling algorithm is not named, described, or validated. Section 5.2 states only that 'we applied an algorithm to label tweets according to their primary topics and sentiment' before reporting that 'cynicism became the dominant sentiment.' No training data, model architecture, prompt template, lexicon, decision threshold, or accuracy measure is given, and the category set mixes attitudes ('cynicism'), topics ('COVID-19 comparisons'), and veracity claims ('misinformation'). Since the paper's main analytical result depends entirely on these labels, the authors should specify the algorithm, report validation against a human-coded gold standard (including inter-annotator agreement or precision/recall/F1), and define each category operationally.
  2. [Abstract and §5.2] The claim of a 'marked increase in tweet volume compared to 2023' is not supported by any quantitative evidence: no raw tweet counts, rates, collection durations, or sample sizes are reported for either year. Moreover, §4 describes a Zeeschuimer collection of 3,000 tweets from 2024 but gives no comparable 2023 collection procedure, so the baseline for the comparison is undefined. The authors should report the underlying numbers and describe how the 2023 and 2024 samples were collected, including any differences in collection mechanism that could confound the volume comparison.
  3. [§4 and Figure 3] The three constituent datasets use different collection mechanisms (Twitter Spritzer Stream random sample versus browser-capture-based Zeeschuimer versus a third-party May 2022 dataset), yet the Methods section does not report the date range, keyword filter, deduplication, or per-dataset tweet counts. As a result, the time series in Figure 3, which begins in April 2024, may reflect collection artifacts rather than genuine shifts in discourse. The authors should state precisely which dataset contributes which dates, how duplicate tweets were handled, and how the daily proportions in Figure 3 were computed, including the denominator.
minor comments (6)
  1. [Title page] The running head contains a typo: 'MPOX-R ELATED' should be 'MPOX-RELATED.'
  2. [References] Reference [11] is cited for the claim that public confusion was intensified by the absence of FDA-approved at-home testing in the United States, but the cited paper addresses COVID-19 vaccination hesitancy in South Africa; the citation appears mismatched and should be corrected or replaced.
  3. [References] References [11] and [14] are the same arXiv preprint (Perikli et al., 2307.15072) listed twice; the duplicate should be removed and the citation numbering adjusted.
  4. [Throughout] The platform is referred to inconsistently as 'X (formerly Twitter),' 'Twitter,' and 'tweets'; the authors should choose one terminology and apply it consistently.
  5. [§5.1 and Figure 3] Figure 3 has no axis labels or y-axis units, so the reader cannot tell what 'proportion' means; the axes and the denominator of the proportion should be labeled clearly.
  6. [§4] The paper reports that Zeeschuimer collected 3,000 tweets from 2024 but does not report the sizes of the other two datasets; reporting all dataset sizes and date ranges would improve reproducibility.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the paper is a descriptive dashboard report, not a fitted prediction or derivation; its unvalidated sentiment-labeling step is an internal-validity concern rather than a construction-level circularity.

full rationale

The paper makes no derivational claim that reduces to its inputs by construction. The abstract's volume comparison ('marked increase in tweet volume compared to 2023') is a descriptive observation, not a quantity fitted from the data and then renamed a prediction. The central finding that 'cynicism became the dominant sentiment' depends on an algorithm that is neither named nor validated in Section 5.2, but this is an unverifiability/internal-consistency problem, not circularity: the label set is not defined in terms of the output conclusion, and no equation equates the input labels to the claimed trend. The only self-citation is to the authors' prior PoxVerifi work ([9]), cited as related prior work on verification ('the potential of AI models to identify credible versus misleading information with over 96 percent accuracy'); it is not used to define the dashboard's design, sentiment labels, or volume comparison, so it is not load-bearing. Because the paper is self-contained as a systems/descriptive report and does not fit parameters to one subset and then predict a closely related quantity, no specific circular step can be quoted, and the score is accordingly low.

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

The paper introduces no new free parameters or invented entities. It rests on a set of unstated data collection and labeling assumptions, particularly the representativeness of the tweet samples and the validity of the sentiment clustering algorithm.

assumptions (4)
  • domain assumption The Spritzer Stream is a representative random sample of approximately 1% of global tweets.
    Stated in Section 4 as an API-based random sample, but no evidence is given for representativeness of mpox-related tweets.
  • domain assumption The 3,000 tweets collected via Zeeschuimer are representative of mpox discourse on X in 2024.
    Zeeschuimer is a browser extension for collecting visible tweets; the sample may be biased by the collectors' search queries and scrolling behavior, which are not described.
  • domain assumption Keyword filtering using 'mpox' and 'Mpox' captures relevant mpox discourse without systematic bias.
    Section 4 states a Python script filters for keywords, but no details on case sensitivity, language coverage, or hashtags are given.
  • ad hoc to paper The unspecified sentiment and topic labeling algorithm produces valid labels for clusters such as cynicism and misinformation.
    Section 5.2 refers to an algorithm without naming it or providing validation; the central finding of cynicism dominance depends on these labels.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Dashboard Approach to Monitoring Mpox-Related Discourse and Misinformation on Social Media." pith.science (2026). https://pith.science/paper/C5R5PHJH

@misc{pith2026250520584,
  author       = {Pith},
  title        = {Pith review of: A Dashboard Approach to Monitoring Mpox-Related Discourse and Misinformation on Social Media},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C5R5PHJH}},
  note         = {Machine review of arXiv:2505.20584}
}
read the original abstract

Mpox (formerly monkeypox) is a zoonotic disease caused by an orthopoxvirus closely related to variola and remains a significant global public health concern. During outbreaks, social media platforms like X (formerly Twitter) can both inform and misinform the public, complicating efforts to convey accurate health information. To support local response efforts, we developed a researcher-focused dashboard for use by public health stakeholders and the public that enables searching and visualizing mpox-related tweets through an interactive interface. Following the CDC's designation of mpox as an emerging virus in August 2024, our dashboard recorded a marked increase in tweet volume compared to 2023, illustrating the rapid spread of health discourse across digital platforms. These findings underscore the continued need for real-time social media monitoring tools to support public health communication and track evolving sentiment and misinformation trends at the local level.

Figures

Figures reproduced from arXiv: 2505.20584 by the authors.

Figure 1
Figure 1. Wireframe mock-up of the Mpox Dashboard interface. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Front page of the mpox dashboard. 5.2 Sentiment and Topic Clustering Over Time To enable comparisons over time, we applied an algorithm to label tweets according to their primary topics and sentiment, then visualized the results using a color-coded cluster graph (see [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Mpox tweet cluster proportions over time (starting April 2024). [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 22 canonical work pages

  1. [1]

    Unveiling the Mpox menace: exploring the intricacies of a zoonotic virus and clinical implications

    Eshita Sharma et al. “Unveiling the Mpox menace: exploring the intricacies of a zoonotic virus and clinical implications”. In: Diagnostic Microbiology and Infectious Disease 107.2 (2023), p. 116024. ISSN : 0732-8893. DOI: https://doi.org/10.1016/j.diagmicrobio.2023.116024. URL: https://www.sciencedirect. com/science/article/pii/S0732889323001347

    work page arXiv 2023

  2. [2]

    World Health Organization. Mpox. Accessed: 2025-05-06. World Health Organization. 2025. URL: https: //www.who.int/news-room/fact-sheets/detail/mpox

    work page 2025

  3. [3]

    Signs and Symptoms of Mpox

    Centers for Disease Control and Prevention. Signs and Symptoms of Mpox. Accessed: 2025-05-06. Centers for Disease Control and Prevention. 2024. URL: https://www.cdc.gov/mpox/signs-symptoms/index.html

    work page 2025

  4. [4]

    The future of infodemic surveillance as public health surveillance

    Howard Chiou et al. “The future of infodemic surveillance as public health surveillance”. In:Emerging Infectious Diseases 28.Suppl 1 (2022), S121

    work page 2022

  5. [5]

    A Prologue to the Special Issue: Health Misinformation on Social Media

    Wen-Ying Sylvia Chou and Anna Gaysynsky. “A Prologue to the Special Issue: Health Misinformation on Social Media”. In: American Journal of Public Health 110.S3 (2020). PMID: 33001727, S270–S272. DOI: 10.2105/AJPH.2020.305943 . eprint: https://doi.org/10.2105/AJPH.2020.305943 . URL: https: //doi.org/10.2105/AJPH.2020.305943

    work page arXiv 2020

  6. [6]

    COVID-19: health literacy is an underestimated problem

    Leena Paakkari and Orkan Okan. “COVID-19: health literacy is an underestimated problem”. In: The lancet public health 5.5 (2020), e249–e250

    work page 2020

  7. [7]

    From belief in conspiracy theories to trust in others: Which factors influence exposure, believing and sharing fake news

    Daniel Halpern et al. “From belief in conspiracy theories to trust in others: Which factors influence exposure, believing and sharing fake news”. In: Social Computing and Social Media. Design, Human Behavior and Analytics: 11th International Conference, SCSM 2019, Held as Part of the 21st HCI International Conference, HCII 2019, Orlando, FL, USA, July 26-...

    work page 2019

  8. [8]

    Contrasting misinformation and real-information dissemination network structures on social media during a health emergency

    Lida Safarnejad et al. “Contrasting misinformation and real-information dissemination network structures on social media during a health emergency”. In: American journal of public health 110.S3 (2020), S340–S347

    work page 2020

Show all 27 references
  1. [9]

    PoxVerifi: An Information Verification System to Combat Monkeypox Misinformation

    Akaash Kolluri, Kami Vinton, and Dhiraj Murthy. PoxVerifi: An Information Verification System to Combat Monkeypox Misinformation. 2022. arXiv: 2209.09300 [cs.CL] . URL: https://arxiv.org/abs/2209. 09300

    2022 arXiv

  2. [10]

    Introduction of Mobile Health Tools to Support Ebola Surveillance and Contact Tracing in Guinea

    Jilian A Sacks et al. “Introduction of Mobile Health Tools to Support Ebola Surveillance and Contact Tracing in Guinea”. In: Global Health: Science and Practice 3.4 (2015), pp. 646–659. DOI: 10 . 9745 / GHSP - D - 15 - 00207. eprint: https : / / www . ghspjournal . org / conte...

    2015

  3. [12]

    Misinformation and Public Health Messaging in the Early Stages of the Mpox Outbreak: Mapping the Twitter Narrative with Deep Learning (Preprint)

    Andy Edinger et al. “Misinformation and Public Health Messaging in the Early Stages of the Mpox Outbreak: Mapping the Twitter Narrative with Deep Learning (Preprint)”. In: Journal of Medical Internet Research 25 (Oct. 2022). DOI: 10.2196/43841

    2022 doi

  4. [13]

    Sentiment analysis and text analysis of the public discourse on Twitter about COVID-19 and MPox

    Nirmalya Thakur. “Sentiment analysis and text analysis of the public discourse on Twitter about COVID-19 and MPox”. In: Big Data and Cognitive Computing 7.2 (2023), p. 116

    2023

  5. [14]

    Detecting the presence of COVID-19 vaccination hesitancy from South African twitter data using machine learning

    Nicholas Perikli et al. “Detecting the presence of COVID-19 vaccination hesitancy from South African twitter data using machine learning”. In: arXiv preprint arXiv:2307.15072 (2023)

    2023 arXiv

  6. [15]

    The production of information in the attention economy

    Giovanni Luca Ciampaglia, Alessandro Flammini, and Filippo Menczer. “The production of information in the attention economy”. In: Scientific reports 5.1 (2015), p. 9452

    2015

  7. [16]

    Correction as a Solution for Health Misinformation on Social Media

    Emily K. Vraga and Leticia Bode. “Correction as a Solution for Health Misinformation on Social Media”. In: American Journal of Public Health 110.S3 (2020). PMID: 33001724, S278–S280. DOI: 10.2105/AJPH.2020. 305916. eprint: https://doi.org/10.2105/AJPH.2020.305916 . URL: https:...

    2020

  8. [17]

    Human-Misinformation interaction: Understanding the interdisciplinary approach needed to computationally combat false information

    Alireza Karduni. Human-Misinformation interaction: Understanding the interdisciplinary approach needed to computationally combat false information. 2019. arXiv: 1903.07136 [cs.HC]. URL: https://arxiv.org/ abs/1903.07136

    2019 arXiv

  9. [18]

    Political science. Exposure to ideologically diverse news and opinion on Facebook

    Eytan Bakshy, Solomon Messing, and Lada Adamic. “Political science. Exposure to ideologically diverse news and opinion on Facebook”. In: Science (New York, N.Y.)348 (May 2015). DOI: 10.1126/science.aaa1160

    2015 doi

  10. [19]

    Understanding the promise and limits of automated fact-checking

    D Graves. “Understanding the promise and limits of automated fact-checking”. In: Reuters Institute for the Study of Journalism (2018). 6

    2018

  11. [20]

    Falling for Fake News: Investigating the Consumption of News via Social Media

    Martin Flintham et al. “Falling for Fake News: Investigating the Consumption of News via Social Media”. In: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems . CHI ’18. Montreal QC, Canada: Association for Computing Machinery, 2018, pp. 1–10. ISBN : ...

    2018

  12. [21]

    RumorLens: Interactive Analysis and Validation of Suspected Rumors on Social Media

    Ran Wang et al. “RumorLens: Interactive Analysis and Validation of Suspected Rumors on Social Media”. In: CHI Conference on Human Factors in Computing Systems Extended Abstracts. ACM, Apr. 2022, pp. 1–7. DOI: 10.1145/3491101.3519712. URL: http://dx.doi.org/10.1145/3491101.3519712

    2022

  13. [22]

    Dashboard of sentiment in Austrian social media during COVID-19

    Max Pellert et al. Dashboard of sentiment in Austrian social media during COVID-19 . June 2020. DOI: 10. 48550/arXiv.2006.11158

    2020 doi arXiv

  14. [23]

    The Development and Psychometric Properties of LIWC-22

    Ryan Boyd et al. The Development and Psychometric Properties of LIWC-22. Feb. 2022. DOI: 10.13140/RG.2. 2.23890.43205

    2022

  15. [24]

    Iannone et al

    J. Iannone et al. flexdashboard: R Markdown format for flexible dashboards. https://pkgs.rstudio.com/ flexdashboard/. R package version 0.6.1, RStudio. 2023

    2023

  16. [25]

    The Johns Hopkins University Center for Systems Science and Engineering COVID-19 Dashboard: data collection process, challenges faced, and lessons learned

    Ensheng Dong et al. “The Johns Hopkins University Center for Systems Science and Engineering COVID-19 Dashboard: data collection process, challenges faced, and lessons learned”. In: The Lancet Infectious Diseases 22 (Aug. 2022). DOI: 10.1016/S1473-3099(22)00434-0

    2022 doi

  17. [26]

    A real-time dashboard of clinical trials for COVID-19

    Kristian Thorlund et al. “A real-time dashboard of clinical trials for COVID-19”. In: The Lancet Digital Health 2 (Apr. 2020). DOI: 10.1016/S2589-7500(20)30086-8

    2020 doi

  18. [27]

    A Twitter dataset for Monkeypox, May 2022

    Zahra M Nia et al. “A Twitter dataset for Monkeypox, May 2022”. In: Data in brief 48 (June 2023), p. 109118. ISSN : 2352-3409. DOI: 10 . 1016 / j . dib . 2023 . 109118. URL: https : / / europepmc . org / articles / PMC10102531

    2022

  19. [28]

    Zeeschuimer

    Digital Methods Initiative. Zeeschuimer. https : / / github . com / orgs / digitalmethodsinitiative / repositories. GitHub repository. 2023. 7

    2023

Pith tools

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