Mapping Discourse Reframing: A Multi-Layer Network Approach to Italian HPV Vaccine Discourse on X (2010-2024)

Lorella Viola

arxiv: 2605.02629 · v1 · submitted 2026-05-04 · 💻 cs.CL · cs.CY· cs.NI

Mapping Discourse Reframing: A Multi-Layer Network Approach to Italian HPV Vaccine Discourse on X (2010-2024)

Lorella Viola This is my paper

Pith reviewed 2026-05-08 19:17 UTC · model grok-4.3

classification 💻 cs.CL cs.CYcs.NI

keywords HPV vaccine discoursemulti-layer networkshashtag co-occurrenceinformation disorderdiscourse reframingsocial media analysispolarizationX platform

0 comments

The pith

Projecting fringe hashtags onto stable core coalitions recovers more long-tail problematic signals in Italian HPV vaccine discourse on X while keeping coalition structures readable.

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

The paper introduces a dual-layer network framework to track how narratives about the HPV vaccine shift and polarize in 14 years of Italian posts on X. It first applies conservative community detection to hashtag co-occurrence networks, isolating a steady prevention-focused core alongside growing skepticism groups. A second coverage layer then maps sparse fringe hashtags into those cores by measuring their weighted links to a set of manually labelled skeptical and conspiratorial seed tweets. This projection step surfaces more problematic long-tail content without erasing the overall coalition pattern. The result is a way to locate where discourse gets reframed and amplified by information disorder over successive time periods.

Core claim

The authors propose a dual-layer approach to hashtag co-occurrence networks. Conservative community detection first identifies a stable prevention-oriented backbone contrasted with increasingly separable skepticism coalitions. A coverage layer then projects fringe hashtags into these core coalitions based on weighted connectivity to a manually labelled set of skeptical and conspiratorial seed tweets. This core-coverage projection significantly improves the recovery of long-tail, problematic hashtags while preserving an interpretable coalition structure, allowing the structural maturation of polarized narratives and the reframing of discourse by information disorder to be mapped over time.

What carries the argument

The core-coverage projection, which maps low-frequency fringe hashtags into robust core coalitions identified by conservative community detection, using weighted connectivity from labelled seed tweets.

If this is right

A stable prevention-oriented backbone persists across the full period while skepticism coalitions become more separable.
The method locates points where online discourse is reframed and amplified by information disorder.
Polarized narratives in health topics show measurable structural maturation over successive epochs.
The dual-layer construction supplies a reusable methodology for tracing discourse reframing in other domains.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same projection technique could be tested on health or political topics outside vaccine discourse to check whether fringe signals integrate in similar patterns.
Early detection of emerging skepticism coalitions via this layer might allow faster public-health monitoring of narrative shifts.
Repeating the analysis on platforms other than X or in languages besides Italian would test whether the dual-layer advantage holds beyond the current dataset.

Load-bearing premise

Manually labelled skeptical and conspiratorial seed tweets supply an unbiased and representative basis for projecting fringe hashtags into core coalitions without selection artifacts.

What would settle it

Applying the projection step to a fresh collection of tweets without using the labelled seed set and observing whether recovery of known problematic long-tail hashtags falls to the level of single-layer methods, or whether the resulting coalition structure becomes harder to interpret.

Figures

Figures reproduced from arXiv: 2605.02629 by Lorella Viola.

**Figure 1.** Figure 1: Hashtag co-occurrence networks for 2010–2014 (coverage vs. core). view at source ↗

**Figure 2.** Figure 2: Hashtag co-occurrence networks for 2015–2019 (coverage vs. core). view at source ↗

**Figure 3.** Figure 3: Hashtag co-occurrence networks for 2020–2024 (coverage vs. core). view at source ↗

**Figure 4.** Figure 4: Augmented assignments by epoch. community detection, while a larger set of hashtags can be projected onto the core community structure based on their connectivity to core nodes. A third group remains unassigned, indicating peripheral or disconnected usage that cannot be reliably linked to a community view at source ↗

**Figure 5.** Figure 5: Core vs coverage enrichment overlay (2020–2024). view at source ↗

**Figure 6.** Figure 6: Distribution of projection support by epoch (sum of edge weights from a projected hashtag to its assigned core community) view at source ↗

read the original abstract

Understanding how online narratives travel through coalitions is critical for identifying information disorder, yet computational analyses often rely on conservative network constructions that erase initially sparse but salient signals. This paper proposes a novel multi-layer framework that captures low-frequency signals of emerging information disorder allowing for locating where online discourse is reframed and amplified over time. The use case is 14 years of Italian discourse on X regarding the Human Papillomavirus (HPV) vaccine across three pivotal epochs (2010-2024). Utilizing hashtag co-occurrence networks, we introduce a dual-layer approach. We first identify robust core discourse coalitions through conservative community detection, revealing a stable prevention-oriented backbone contrasted with increasingly separable skepticism coalitions. We then introduce a coverage layer and project fringe hashtags into core coalitions based on weighted connectivity. Using a manually labelled set of skeptical and conspiratorial seed tweets, we demonstrate that this core-coverage projection significantly improves the recovery of long-tail, problematic hashtags while preserving an interpretable coalition structure. Our findings characterize the structural maturation of polarized narratives and provide a methodology for mapping how discourse is reframed and amplified by information disorder over time.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The dual-layer core-coverage method recovers more long-tail hashtags in the Italian HPV discourse network but the improvement claim rests on unvalidated manual seeds.

read the letter

The paper introduces a dual-layer network method for hashtag co-occurrence data: first run conservative community detection to find stable core coalitions, then add a coverage layer that projects fringe hashtags into those cores using weighted connectivity to a set of manually labelled skeptical and conspiratorial seed tweets. The case is 14 years of Italian X posts on the HPV vaccine, split across three epochs from 2010-2024. It shows a persistent prevention-oriented backbone alongside increasingly separable skepticism coalitions and claims the projection step pulls in more problematic long-tail hashtags without breaking interpretability. That is the actual novelty on offer. The application to a longitudinal national dataset is straightforward and the structural findings on narrative maturation are clear enough to be useful for people tracking health discourse online. The soft spot is exactly where the stress-test note points: the abstract states a significant improvement in recovery but supplies no numbers, no error analysis, no inter-annotator agreement, and no sampling details on the seed tweets. Without those checks it is difficult to know whether the gain is structural or an artifact of how the seeds were chosen. If the full paper has those diagnostics and they hold up, the method becomes more convincing; right now the claim is under-supported. This is for computational social scientists working on misinformation, network methods for sparse text data, or vaccine discourse specifically. A reader who needs a worked example of handling long-tail signals in social media networks will get something concrete from it. The work is coherent on its own terms and shows honest engagement with the problem of conservative network constructions, so it deserves a serious referee even if the validation section will need strengthening.

Referee Report

2 major / 1 minor

Summary. The paper proposes a multi-layer network framework for mapping discourse reframing in 14 years of Italian HPV vaccine discussions on X (2010-2024). It constructs hashtag co-occurrence networks, applies conservative community detection to identify stable core coalitions (prevention-oriented backbone vs. separable skepticism coalitions), and introduces a coverage layer that projects fringe hashtags into core coalitions via weighted connectivity to a manually labelled set of skeptical/conspiratorial seed tweets. The central claim is that this projection significantly improves recovery of long-tail problematic hashtags while preserving interpretable coalition structure, enabling characterization of polarized narrative maturation and information disorder.

Significance. If the projection method is validated, the work offers a useful approach for capturing low-frequency signals in online health discourse without erasing sparse but salient elements, with the 14-year multi-epoch scope and dual-layer design as clear strengths. It contributes to computational analyses of coalition formation in misinformation contexts. Credit for grounding the coverage layer in external manual labels rather than purely internal network properties.

major comments (2)

[Abstract] Abstract: The assertion that the core-coverage projection 'significantly improves the recovery of long-tail, problematic hashtags' is presented without any quantitative metrics, baseline comparisons, error analysis, or validation statistics. This is load-bearing for the central methodological claim.
[Methods] Manual labelling procedure: The projection step anchors on a 'manually labelled set of skeptical and conspiratorial seed tweets,' yet no details are supplied on labelling criteria, sample size, inter-annotator agreement, sampling strategy, or bias mitigation. This directly affects whether the reported improvement reflects genuine structural properties or selection artifacts in the seeds.

minor comments (1)

[Abstract] Abstract: Adding one or two concrete performance numbers (e.g., recall improvement or F1 scores) would strengthen the summary of results.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback. We address the two major comments point by point below. Both points identify areas where the current manuscript is insufficiently explicit, and we have revised the text accordingly to strengthen the presentation of the core-coverage projection and its grounding.

read point-by-point responses

Referee: [Abstract] Abstract: The assertion that the core-coverage projection 'significantly improves the recovery of long-tail, problematic hashtags' is presented without any quantitative metrics, baseline comparisons, error analysis, or validation statistics. This is load-bearing for the central methodological claim.

Authors: We agree that the abstract, standing alone, does not supply the quantitative support required for this central claim. The Results section of the manuscript reports the relevant evaluation, including direct comparisons against the core-only network (showing higher recall of long-tail skeptical hashtags while coalition modularity remains stable) together with precision estimates derived from the seed set. In the revised version we will condense the key metrics, baseline contrast, and validation approach into the abstract so that the claim is self-contained and substantiated. revision: yes
Referee: [Methods] Manual labelling procedure: The projection step anchors on a 'manually labelled set of skeptical and conspiratorial seed tweets,' yet no details are supplied on labelling criteria, sample size, inter-annotator agreement, sampling strategy, or bias mitigation. This directly affects whether the reported improvement reflects genuine structural properties or selection artifacts in the seeds.

Authors: We accept that the current Methods section omits the necessary procedural details. The revised manuscript will add a dedicated subsection that specifies: (i) the annotation guidelines and criteria (drawing on established misinformation taxonomies), (ii) the exact sample size and sampling frame, (iii) inter-annotator agreement statistics, and (iv) bias-mitigation steps (multiple annotators, adjudication protocol, and documentation of disagreement resolution). These additions will make transparent that the seed set is externally grounded rather than an artifact of internal network properties. revision: yes

Circularity Check

0 steps flagged

No significant circularity: derivation grounded in external manual labels

full rationale

The paper constructs core coalitions via conservative community detection on hashtag co-occurrence networks, then projects fringe hashtags via weighted connectivity to a manually labelled set of skeptical/conspiratorial seed tweets. This manual labelling step supplies an independent external anchor rather than deriving the projection or its claimed improvement from the network structure alone. No self-definitional equations, fitted parameters renamed as predictions, load-bearing self-citations, or ansatz smuggling appear in the described method. The improvement claim is evaluated against the external seeds, keeping the chain non-circular.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit free parameters, axioms, or invented entities; the approach extends standard hashtag co-occurrence and community detection without introducing new postulated constructs.

pith-pipeline@v0.9.0 · 5498 in / 1042 out tokens · 57370 ms · 2026-05-08T19:17:25.673354+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages · 1 internal anchor

[1]

Introduction Discourses rise and fall in popularity not arbitrar- ily, but because they align with the prevailing so- cial, political, and economic contexts in which they are created and perpetuated (Robertson, 1998). For example, public conceptualisations and discus- sions of health reflect more than just health: they also offer privileged viewpoints to ...

work page internal anchor Pith review Pith/arXiv arXiv 1998
[2]

This work high- lights the dynamic interplay between public priori- ties, political decision-making, and the framing of responsibilityandrisk

Previous studies Research on health discourse has long empha- sized its temporal dimension, showing how shifts in public sentiment and dominant narratives can shape, sometimes rapidly, policy agendas and in- stitutional responses (McPhail and Bombak, 2015; Berridge, 2007; Lawrence, 2004). This work high- lights the dynamic interplay between public priori-...

work page 2015
[3]

Ofthe 100 types of HPV, 13 may induce cancer, with cer- vical cancer being the most common HPV-related cancer and the fourth most common cancer among women worldwide (WHO, 2022)

The HPV vaccine HPV, the most prevalent sexually transmitted infec- tion globally, affects both men and women, with an estimated80%-90%ofindividualscontractingitdur- ingtheirlifetime(KombeKombeetal.,2021). Ofthe 100 types of HPV, 13 may induce cancer, with cer- vical cancer being the most common HPV-related cancer and the fourth most common cancer among w...

work page 2021
[4]

Data and methodology Data retrieval was conducted through targeted queries that extracted posts from X containing spe- cific hashtags, including #HPV, #Gardasil, #papil- lomavirus, #papilloma.1 Even though posts were collected with the platform language parameter set to Italian (lang=it), an additional post-hoc language filter was necessary because the pl...

work page 2010
[5]

Coverage networks are larger and typically in- clude more peripheral components, while the core networksconcentratemassintofewer,denserstruc- tures

Analysis and Results First, we report the size of thelargest connected component(LCC), i.e., the number of nodes in the largest subgraph in which all nodes are mutually reachable via paths, together with the total number of disconnected components in the network (implic- itly). Coverage networks are larger and typically in- clude more peripheral component...

work page 2010
[6]

Validation To validate the methodology, we connected hashtag communities to higher-level information- disorder narratives by constructing a small seed lexicon of narrative-indicative hashtags for two frames:skeptical/anti-vaccineandconspiracy. The skeptical/anti-vaccine frame included posts ex- pressing doubt or opposition to vaccination, such assafety/ef...

work page 2020
[7]

In the three epochs, clear core clus- ters shifts can be identified: institutional discourse forms the core in 2010-2014 and skepticism is sparse and located in the long tail

Discussion The 14-year longitudinal analysis (2010–2024) pro- vides empirical evidence for the lifecycle of HPV vaccine discourse in Italian X through narrative coalitions. In the three epochs, clear core clus- ters shifts can be identified: institutional discourse forms the core in 2010-2014 and skepticism is sparse and located in the long tail. In the s...

work page 2010
[8]

Conclusion This paper examined the temporal evolution of Ital- ian discourse about the HPV vaccine on X through a graph-based analysis of hashtag co-occurrence. By modelling hashtags as cultural objects that mark discourse and tracking their co-occurrence structure across three epochs (2010-2014, 2015- 2019, 2020-2024), cross-epoch community match- ing, a...

work page 2010
[9]

Limitations Whilethisstudyprovideseveralinsights, limitations must be acknowledged. First, the dataset is sub- ject to the X’s API access constraints and policy requirements and observed temporal differences may partially reflect changes in platform usage and data availability rather than discourse alone. Sec- ond, hashtag-based analysis captures a salien...

work page
[10]

2007.Marketing health: smok- ing and the discourse of public health in Britain, 1945-2000

References Virginia Berridge. 2007.Marketing health: smok- ing and the discourse of public health in Britain, 1945-2000. Oxford University Press, Oxford ; New York. OCLC: ocn123797267. Erika Bonnevie, Allison Gallegos-Jeffrey, Jaclyn Goldbarg, Brian Byrd, and Joseph Smyser. 2021. Quantifying the rise of vaccine opposition on Twitter during the COVID-19 pa...

work page doi:10.1080/10410236.2011.610258 2007
[11]

Diana Dobrin

Publisher: Taylor & Francis. Diana Dobrin. 2020. The Hashtag in Digital Ac- tivism: A Cultural Revolution.Journal of Cultural Analysis and Social Change, 5(1):03. Nihal Durmaz and Engin Hengirmen. 2022. The dramatic increase in anti-vaccine discourses during the COVID-19 pandemic: a social network analysis of Twitter.Human Vac- cines & Immunotherapeutics,...

work page doi:10.1080/21645515.2021.2025008 2020
[12]

Fat, queer and sick? A critical anal- ysis of ‘lesbian obesity’ in public health discourse.Critical Public Health, 25(5):539–

work page
[13]

Francesco Saverio Mennini, Andrea Silenzi, An- drea Marcellusi, Michele Conversano, Andrea Siddu, and Giovanni Rezza

Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/09581596.2014.992391. Francesco Saverio Mennini, Andrea Silenzi, An- drea Marcellusi, Michele Conversano, Andrea Siddu, and Giovanni Rezza. 2022. HPV Vacci- nation during the COVID-19 Pandemic in Italy: Opportunity Loss or Incremental Cost.Vaccines, 10(7):1133. Bjarke Mønsted and Sune Lehmann. 2...

work page doi:10.1080/09581596.2014.992391 2014
[14]

Michele Zappavigna

HPV Vaccine Issues in Japan: A review of our attempts to promote the HPV vaccine and to provide effective evaluation of the problem through social-medical and behavioral-economic perspectives.Vaccine, 42(22):125859. Michele Zappavigna. 2016. Twitter. In Christian R. HoffmannandWolframBublitz,editors,Pragmat- ics of social media, volume 11, pages 201–224. ...

work page 2016

[1] [1]

Introduction Discourses rise and fall in popularity not arbitrar- ily, but because they align with the prevailing so- cial, political, and economic contexts in which they are created and perpetuated (Robertson, 1998). For example, public conceptualisations and discus- sions of health reflect more than just health: they also offer privileged viewpoints to ...

work page internal anchor Pith review Pith/arXiv arXiv 1998

[2] [2]

This work high- lights the dynamic interplay between public priori- ties, political decision-making, and the framing of responsibilityandrisk

Previous studies Research on health discourse has long empha- sized its temporal dimension, showing how shifts in public sentiment and dominant narratives can shape, sometimes rapidly, policy agendas and in- stitutional responses (McPhail and Bombak, 2015; Berridge, 2007; Lawrence, 2004). This work high- lights the dynamic interplay between public priori-...

work page 2015

[3] [3]

Ofthe 100 types of HPV, 13 may induce cancer, with cer- vical cancer being the most common HPV-related cancer and the fourth most common cancer among women worldwide (WHO, 2022)

The HPV vaccine HPV, the most prevalent sexually transmitted infec- tion globally, affects both men and women, with an estimated80%-90%ofindividualscontractingitdur- ingtheirlifetime(KombeKombeetal.,2021). Ofthe 100 types of HPV, 13 may induce cancer, with cer- vical cancer being the most common HPV-related cancer and the fourth most common cancer among w...

work page 2021

[4] [4]

Data and methodology Data retrieval was conducted through targeted queries that extracted posts from X containing spe- cific hashtags, including #HPV, #Gardasil, #papil- lomavirus, #papilloma.1 Even though posts were collected with the platform language parameter set to Italian (lang=it), an additional post-hoc language filter was necessary because the pl...

work page 2010

[5] [5]

Coverage networks are larger and typically in- clude more peripheral components, while the core networksconcentratemassintofewer,denserstruc- tures

Analysis and Results First, we report the size of thelargest connected component(LCC), i.e., the number of nodes in the largest subgraph in which all nodes are mutually reachable via paths, together with the total number of disconnected components in the network (implic- itly). Coverage networks are larger and typically in- clude more peripheral component...

work page 2010

[6] [6]

Validation To validate the methodology, we connected hashtag communities to higher-level information- disorder narratives by constructing a small seed lexicon of narrative-indicative hashtags for two frames:skeptical/anti-vaccineandconspiracy. The skeptical/anti-vaccine frame included posts ex- pressing doubt or opposition to vaccination, such assafety/ef...

work page 2020

[7] [7]

In the three epochs, clear core clus- ters shifts can be identified: institutional discourse forms the core in 2010-2014 and skepticism is sparse and located in the long tail

Discussion The 14-year longitudinal analysis (2010–2024) pro- vides empirical evidence for the lifecycle of HPV vaccine discourse in Italian X through narrative coalitions. In the three epochs, clear core clus- ters shifts can be identified: institutional discourse forms the core in 2010-2014 and skepticism is sparse and located in the long tail. In the s...

work page 2010

[8] [8]

Conclusion This paper examined the temporal evolution of Ital- ian discourse about the HPV vaccine on X through a graph-based analysis of hashtag co-occurrence. By modelling hashtags as cultural objects that mark discourse and tracking their co-occurrence structure across three epochs (2010-2014, 2015- 2019, 2020-2024), cross-epoch community match- ing, a...

work page 2010

[9] [9]

Limitations Whilethisstudyprovideseveralinsights, limitations must be acknowledged. First, the dataset is sub- ject to the X’s API access constraints and policy requirements and observed temporal differences may partially reflect changes in platform usage and data availability rather than discourse alone. Sec- ond, hashtag-based analysis captures a salien...

work page

[10] [10]

2007.Marketing health: smok- ing and the discourse of public health in Britain, 1945-2000

References Virginia Berridge. 2007.Marketing health: smok- ing and the discourse of public health in Britain, 1945-2000. Oxford University Press, Oxford ; New York. OCLC: ocn123797267. Erika Bonnevie, Allison Gallegos-Jeffrey, Jaclyn Goldbarg, Brian Byrd, and Joseph Smyser. 2021. Quantifying the rise of vaccine opposition on Twitter during the COVID-19 pa...

work page doi:10.1080/10410236.2011.610258 2007

[11] [11]

Diana Dobrin

Publisher: Taylor & Francis. Diana Dobrin. 2020. The Hashtag in Digital Ac- tivism: A Cultural Revolution.Journal of Cultural Analysis and Social Change, 5(1):03. Nihal Durmaz and Engin Hengirmen. 2022. The dramatic increase in anti-vaccine discourses during the COVID-19 pandemic: a social network analysis of Twitter.Human Vac- cines & Immunotherapeutics,...

work page doi:10.1080/21645515.2021.2025008 2020

[12] [12]

Fat, queer and sick? A critical anal- ysis of ‘lesbian obesity’ in public health discourse.Critical Public Health, 25(5):539–

work page

[13] [13]

Francesco Saverio Mennini, Andrea Silenzi, An- drea Marcellusi, Michele Conversano, Andrea Siddu, and Giovanni Rezza

Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/09581596.2014.992391. Francesco Saverio Mennini, Andrea Silenzi, An- drea Marcellusi, Michele Conversano, Andrea Siddu, and Giovanni Rezza. 2022. HPV Vacci- nation during the COVID-19 Pandemic in Italy: Opportunity Loss or Incremental Cost.Vaccines, 10(7):1133. Bjarke Mønsted and Sune Lehmann. 2...

work page doi:10.1080/09581596.2014.992391 2014

[14] [14]

Michele Zappavigna

HPV Vaccine Issues in Japan: A review of our attempts to promote the HPV vaccine and to provide effective evaluation of the problem through social-medical and behavioral-economic perspectives.Vaccine, 42(22):125859. Michele Zappavigna. 2016. Twitter. In Christian R. HoffmannandWolframBublitz,editors,Pragmat- ics of social media, volume 11, pages 201–224. ...

work page 2016