CoChat
Sign Up
Cited by 51Social Media and PoliticsMisinformation and Its ImpactsComplex Network Analysis Techniques

Influence of augmented humans in online interactions during voting events

Open access
Stella, Massimo, Cristoforetti, Marco, De Domenico, Manlio·2019·PLOS ONE
DOI 10.1371/journal.pone.0214210PMC PMC6521989
Open this paper in CoChat — free account

Abstract

The advent of the digital era provided a fertile ground for the development of virtual societies, complex systems influencing real-world dynamics. Understanding online human behavior and its relevance beyond the digital boundaries is still an open challenge. Here we show that online social interactions during a massive voting event can be used to build an accurate map of real-world political parties and electoral ranks for Italian elections in 2018. We provide evidence that information flow and collective attention are often driven by a special class of highly influential users, that we name "augmented humans", who exploit thousands of automated agents, also known as bots, for enhancing their online influence. We show that augmented humans generate deep information cascades, to the same extent of news media and other broadcasters, while they uniformly infiltrate across the full range of identified groups. Digital augmentation represents the cyber-physical counterpart of the human desire to acquire power within social systems.

AI snapshot

Written from the paper’s full text.

TL;DR

Online Twitter interactions during the 2018 Italian elections reflected real-world political parties and electoral ranks, with interaction volume correlating strongly with vote share. A class of “augmented humans” used bots to gain influence comparable to media and political broadcasters, producing deep information cascades and infiltrating across social groups.

Key findings

  • The social bulk identified 8 main communities: five mapped to Italian political ecosystems, two to news media, and one to a single augmented human plus bots; group interaction volume ranked with election outcome (Spearman ρ = 0.9, p = 0.039).
  • Bot classification labeled 13.4% of users as bots; humans interacted with bots in 46% of social interactions overall (56% of mentions, 41% of replies, 43% of retweets), while bots interacted with humans in roughly 4%; bots contributed 6% of total social interactions, and 9 out of 10 hubs were bots.
  • The social bulk was polarized (ΦF = 0.452), fragmented (ΦL = 0.812), and segregated (Σ = 0.476 vs random expectation 0.172, p < 10−5). Excluding the augmented-human community (97.9% bots), mean bot infiltration was 29.2% and mean augmented-human infiltration was 15.7%.
  • Augmented humans were defined as human users with at least 50% bot neighbors in the social bulk; 1,010 accounts (12.7% of humans in the social bulk) were identified. The most central, User01, interacted with 2,700 bots and 55 humans; 67% of augmented humans were influentials, hidden influentials, or broadcasters. Cascades numbered 83,593, with sizes 2–4,313 and a power-law exponent γ = −2.33 ± 0.04.

Methods

  • Collected 966,483 tweets from 194,273 users between 24 February and 7 March 2018 using Twitter’s streaming API, filtered by election-related keywords/hashtags; no Twitter warnings were received, so the authors report collecting 100% of tweets containing the specified keywords.
  • Classified users as humans or bots with a deep neural network using 10 account features, trained and validated on 22,993 manually annotated accounts (14,218 bots, 8,775 humans) with an 80/20 split; random forest and deep neural network achieved >90% accuracy and >95% bot precision.
  • Built interaction networks from retweets, mentions, and replies; constructed a “social bulk” by requiring at least one retweet plus either a mention or reply between user pairs. Analyzed polarization via Fiedler partitioning and modularity, fragmentation via Louvain modularity, segregation via connected components, infiltration by user class, and information cascades via retweet counts and cascade rate.

Limitations

  • The paper states that Twitter metadata does not allow full reconstruction of information cascades because of missing intermediary retweets, so cascades were measured as star networks by total retweets received. It also states that further research is needed to generalize the correlation between online interaction volume and electoral outcomes and confirm predictive power.
  • Evident from the text: the study is an observational, single-country, single-election case study, so causal claims about bots or augmented humans influencing real-world voting are not established. Bot and augmented-human classifications depend on the trained model and thresholds (e.g., augmented humans defined as ≥50% bot neighbors, users with fewer than 3 bulk interactions discarded), and the data are limited to keyword-filtered tweets collected through Twitter’s streaming API.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0214210

Cites

52 of 52 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 31 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    Online social interactions during a massive voting event can be used to build an accurate map of real-world political parties and electoral ranks for Italian elections in 2018.

    “Here we show that online social interactions during a massive voting event can be used to build an accurate map of real-world political parties and electoral ranks for Italian elections in 2018.”
    Quote in the paperSubstantial claim
  • Findingown data

    Information flow and collective attention are often driven by a special class of highly influential users, named 'augmented humans', who exploit thousands of automated agents (bots) for enhancing their online influence.

    “We provide evidence that information flow and collective attention are often driven by a special class of highly influential users, that we name “augmented humans”, who exploit thousands of automated agents, also known as bots, for enhancing their online influence.”
    Quote in the paperSubstantial claim
  • Findingown data

    Augmented humans generate deep information cascades, to the same extent as news media and other broadcasters.

    “We show that augmented humans generate deep information cascades, to the same extent of news media and other broadcasters, while they uniformly infiltrate across the full range of identified groups.”
    Quote in the paperSubstantial claim
  • Findingnone given

    Augmented humans uniformly infiltrate across the full range of identified groups.

    “We show that augmented humans generate deep information cascades, to the same extent of news media and other broadcasters, while they uniformly infiltrate across the full range of identified groups.”
    Quote in the paperSubstantial claim
  • Findingown data

    Bots interact with humans roughly in 4% of the interactions, independently on interaction type.

    “Bots interact with humans roughly in 4% of the interactions, independently on interaction type.”
    Quote in the paperSubstantial claim
  • May combine two claims. 33
  • Classified as probably not a substantial claim. 6

  • May combine two claims.

    Findingown data

    Most users are located in Italy, but there are significant fractions of human users also located in the United States and in Europe.

    “Although most of the users are located in Italy, there are significant fractions of human users also located in the United States and in Europe”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The social fragmentation of the original system is stable across all phases of voting and significantly larger than random expectation, confirming that results from the social bulk are consistent.

    “Once again, we observe that social fragmentation is stable across time and significantly larger than random expectation, confirming that results obtained from the social bulk are consistent.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation

    The deep neural network model classified 13.4% of users as bots, a value compatible to estimations during other voting events.

    “To identify automated agents in the data set, we developed a deep neural network model (see Methods and SI), which classified 13.4% of users as bots, a value compatible to estimations during other voting events [10, 11, 12].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [10] E Ferrara, 2016, The rise of social bots, Comm of the ACM source

      Not checked in the free run.

    • cites [11] E Ferrara, 2017, Disinformation and social bot operations in the run up to the 2017 French presidential election, First Monday source

      Not checked in the free run.

    • cites [12] M Stella, 2018, Bots increase exposure to negative and inflammatory content in online social systems, Proceedings of the National Academy of Sciences

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    Humans tend to interact with bots in 56% of mentions, 41% of replies and 43% of retweets.

    “Humans tend to interact with bots in 56% of mentions, 41% of replies and 43% of retweets.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    9 out of 10 hubs-i.e., highly interacting users-are bots and they are mainly news media and public profiles of politicians.

    “In fact, 9 out of 10 hubs—i.e., highly interacting users—are bots and they are mainly news media and public profiles of politicians”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    During Italian elections, bots are mostly broadcasters (mostly media) and influentials (mostly political leaders).

    “Remarkably, the topological and dynamical behavior of humans and bots is very different: during Italian elections, bots are mostly broadcasters (mostly media) and influentials (mostly political leaders).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The segregation of the social bulk is Σ = 0.476, significantly stronger than the random expectation 〈Σrand〉 = 0.172 (p-value < 10−5) from a configuration model preserving the connectivity distribution.

    “The segregation of the social bulk is Σ = 0.476, significantly stronger than random expectation 〈Σrand〉 = 0.172 (p-value < 10−5) based on a configuration model preserving the connectivity distribution.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Deeper information cascades are generated mostly by humans with a high number of followers, with the remarkable exception of one, User01, who produced the largest cascade among humans and bots despite having less than 100 followers.

    “Deeper information cascades are generated mostly by humans with a high number of followers, with the remarkable exception of one, User01, who produced the largest cascade among humans and bots despite having less than 100 followers.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Excluding the augmented human's community (97.9% bots), the mean bot infiltration in the bulk network is 29.2% and the mean augmented-human infiltration is 15.7%.

    “Excluding the community corresponding to the augmented human (made for 97.9% of bots), the mean bot infiltration in the bulk network is 29.2% while the mean augmented infiltration is 15.7%.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Hub analysis of the social bulk groups shows that one group corresponds to a single augmented human and his/her bots, five groups map the ecosystems of the main Italian political parties, and two groups encode the news media universe (traditional or online).

    “The analysis of hubs in each group of the social bulk indicates that i) one group corresponds to a single augmented human and his/her bots; ii) five groups directly map the ecosystems of the main Italian political parties; iii) two groups encode news media universe, either traditional or online news organisations.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    At night the volume of messages generated by humans drops down (circadian rhythm).

    “The circadian rhythm is evident, i.e. at night the volume of messages generated by humans drops down.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Bots display a circadian rhythm similar to humans.

    “Also bots display a similar circadian rhythm, in agreement with previous observations [10, 12].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [10] E Ferrara, 2016, The rise of social bots, Comm of the ACM source

      Not checked in the free run.

    • cites [12] M Stella, 2018, Bots increase exposure to negative and inflammatory content in online social systems, Proceedings of the National Academy of Sciences

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    The distribution of observed cascade sizes is heavy-tailed and compatible with a power law characterized by a scaling exponent γ = −2.33±0.04.

    “As expected for complex networks with highly heterogeneous connectivity [27], the distribution of observed cascade sizes is heavy-tailed and compatible with a power law characterized by a scaling exponent γ = −2.33±0.04, similarly to size distribution in percolation theory or avalanches in self-organized criticality [32].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [27] DJ Watts, 2002, A simple model of global cascades on random networks, Proceedings of the National Academy of Sciences source

      Not checked in the free run.

    • cites [32] SN Dorogovtsev, 2008, Critical phenomena in complex networks, Rev Mod Phys source

      Not checked in the free run.

  • May combine two claims.

    Findingown data

    The augmented human's group consists of more than 2,500 automated agents interacting with the augmented human user, producing a star-like structure for that community (network C8).

    “First, we focused our attention on the augmented human’s group, consisting of more than 2,500 automated agents artificially interacting with the augmented human user. This peculiar activity leads to a star-like structure for the corresponding community, as shown in Fig 3c, network C8.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Bots act as broadcasters by repeating the same political content of human users, boosting the spread of hashtags related to the electoral process.

    “The analysis of topic frequency and associations in bot-generated messages confirms this trend: Bots act as broadcasters by repeating the same political content of human users, boosting the spread of hashtags related to the electoral process (see SI).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The models based on random forest and deep neural network provided the highest accuracy (>90%) and precision in identifying bots (>95%).

    “The models based on random forest and deep neural network provided us with the highest accuracy (>90%) and precision in identifying bots (>95%).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In the whole social bulk, the measured fragmentation is ΦL = 0.812, significantly larger than the null model expectation (p-value < 10−5), indicating several factions in stronger opposition than expected.

    “In the whole social bulk we measure a fragmentation ΦL = 0.812, indicating the presence of several factions in the bulk network that are in a stronger opposition when compared to the null model (, p-value < 10−5).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingnone given

    Bots' locations are distributed worldwide and they are present in areas where no human users are geo-localized such as Morocco, Peru, Finland or Indonesia.

    “Similarly, bots’ locations are distributed worldwide and they are present in areas where no human users are geo-localized such as Morocco, Peru, Finland or Indonesia.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    All the augmented humans identified in this study have, on average, less than 9,000 followers and 1,500 friends.

    “All the augmented humans identified in this study have, on average, less than 9,000 followers and 1,500 friends, indicating that a considerable amount of social influence was obtained by users that preferentially interacted with bots during the considered event.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Removing all users in the two news media communities led to negligible fluctuations (around 0.02) in the fractions of human-bot interactions and in the total volume of tweets produced by bots (around 0.4%).

    “The removal of news media accounts led to negligible fluctuations (around 0.02) in the fractions of human-bot interactions (cfr. Fig 1(a)) and in the total volume of tweets produced by bots (around 0.4%).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation

    Hidden influentials are mostly humans, but augmented humans also fall in this category (e.g. User01).

    “Hidden influentials, known to be efficient spreaders in viral phenomena [44], are mostly humans but augmented humans also falls in this category (e.g. User01).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [44] RA Baños, 2013, The role of hidden influentials in the diffusion of online information cascades, EPJ Data Science source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Many large-scale studies have reported evidence of online ecosystems altering decision-making of crowds and influencing real-world voting of millions of people.

    “Many large-scale studies have reported evidence of online ecosystems altering decision-making of crowds [3] and influencing real-world voting of millions of people [4].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [3] L Muchnik, 2013, Social influence bias: A randomized experiment, Science source

      Not checked in the free run.

    • cites [4] RM Bond, 2012, A 61-million-person experiment in social influence and political mobilization, Nature source

      Not checked in the free run.

  • May combine two claims.

    Study design and conductreasoning

    In the case of Italian elections, no warnings were received, therefore 100% of tweets containing the specified keywords were collected.

    “In the case of Italian elections, we received no warnings, therefore we have collected 100% of tweets containing the specified keywords.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Strong interactions lead to more segregated components, with fewer bridges among connected components than expected from the heterogeneity of interactions alone.

    “This indicates that strong interactions lead to more segregated components, with fewer bridges among connected components than expected from the heterogeneity of interactions only.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The heterogeneity of social interactions can not explain, alone, the observed level of polarization, which has rather to be attributed to other causes such as political parties or opposing political trends.

    “This result indicates that the heterogeneity of social interactions can not explain, alone, the observed level of polarization, which has rather to be attributed to other causes such as political parties or opposing political trends.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: If two users share similar political ideologies, they can endorse and subsequently share (i.e. retweet) the content of each other.

  • May combine two claims.

    Interpretationown data

    Because ΦL > ΦF, the mesoscale organization of the social bulk is more accurately described when more than two groups are considered, consistent with the expectation that results reflect the real-world socio-political scenario.

    “The fact that ΦL > ΦF indicates that a more accurate description of the mesoscale organization of the social bulk is given when more than two groups are considered, in agreement with our hypothesis that results should reflect the real world socio-political scenario.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Retweets are traditionally considered as a form of social endorsement: Users tend to retweet and thus endorse content they agree with.

    “While Mentions and Replies can have both negative and positive connotations, Retweets are traditionally considered as a form of social endorsement [26, 17]: Users tend to retweet and thus endorse content they agree with.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [26] Metaxas PT, Mustafaraj E, Wong K, Zeng L, O’Keefe M, Finn S. What Do Retweets Indicate? Results from User Survey and Meta-Review of Research. In: ICWSM; 2015. p. 658–661. source

      Not checked in the free run.

    • cites [17] S Aral, 2012, Identifying influential and susceptible members of social networks, Science source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    Bots play a passive role in the network but are rather highly targeted by humans.

    “This indicates that bots play a passive role in the network but are rather highly targeted by humans.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    By construction, information flows among users who share strong social relationships and are characterized by similar ideologies.

    “By construction, information flows among users who share strong social relationships and are characterized by similar ideologies: In fact, when a retweet goes from one user to another one, both of them are endorsing the same content, thus making non-directionality a viable approach for representing the endorsement related to content sharing.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: If two users share similar political ideologies, they can endorse and subsequently share (i.e. retweet) the content of each other.

  • May combine two claims.

    Interpretationreasoning

    Bots share messages mainly with each other and hence can resonate with the same content, be it news or spam.

    “Since interactions encode content spread [16], this result indicates that bots share messages mainly with each other and hence can resonate with the same content, be it news or spam.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [16] S Aral, 2009, Distinguishing influence-based contagion from homophily-driven diffusion in dynamic networks, Proceedings of the National Academy of Sciences source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    News media and public profiles of politicians usually act as broadcasters and drive online information flow.

    “which usually act as broadcasters and drive online information flow [17, 15].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [17] S Aral, 2012, Identifying influential and susceptible members of social networks, Science source

      Not checked in the free run.

    • cites [15] S González-Bailón, 2013, Broadcasters and hidden influentials in online protest diffusion, American Behavioral Scientist source

      Not checked in the free run.

  • May combine two claims.

    Study design and conductown data

    The correlation is statistically significant at the 5% level, and the p-value was computed by direct sampling of rankings without relying on assumptions about the large-scale statistical properties of the data.

    “This correlation is statistically significant within a 5% significance level and direct sampling of rankings was used in order to compute the p-value without relying on any assumption about the large-scale statistical properties of the data.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    For this case study, the volume of online interactions closely mirrors the election outcome, though generalizing this result and confirming its predictive power requires further research.

    “The strong correlation found indicates that the volume of online interactions closely mirrors the election outcome for this case study, although further research is needed to generalize this result and confirm its predictive power.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    A prominent amount of human-bot interactions does not involve news media accounts and is not influenced by the presence of information hubs.

    “These results indicate that a prominent amount of human-bot interactions does not involve news media accounts and it is not influenced by the presence of information hubs.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Digital augmentation is a new phenomenon, identified and quantified in this work, characterizing individuals that coordinate from hundreds to thousands of social bots to achieve social influence comparable to that of political parties and news media organisations, with serious real-world repercussions.

    “In this work we have identified and quantified a new phenomenon, i.e. digital augmentation, to characterize individuals that coordinate from hundreds to thousands of social bots for achieving a social influence comparable to the one of political parties and news media organisations, with serious repercussions in the real-world.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationnone given

    This work provides a first step towards a more systematic quantification of the impact of digital augmentation in opinion formation and the manipulation of online attention by means of human-bot interactions.

    “Our work provides a first step towards a more systematic quantification of the impact of digital augmentation in opinion formation and the manipulation of online attention by means of human-bot interactions.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Online ecosystems acted upon society by altering the emotions and beliefs of large populations of individuals.

    “This finding is in full agreement with previous works showing how online ecosystems acted upon society by altering the emotions [19] and beliefs [3, 4] of large populations of individuals.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [19] M Bruter, 2017, Understanding the emotional act of voting, Nature Human Behaviour

      Not checked in the free run.

    • cites [3] L Muchnik, 2013, Social influence bias: A randomized experiment, Science source

      Not checked in the free run.

    • cites [4] RM Bond, 2012, A 61-million-person experiment in social influence and political mobilization, Nature source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    An individual is much more likely to adopt a content when his/her neighbors in the social network tend to reinforce it.

    “The analysis of information and behavioral spreading on social media [17] revealed that an individual is much more likely to adopt a content when his/her neighbors in the social network tend to reinforce it [48].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [17] S Aral, 2012, Identifying influential and susceptible members of social networks, Science source

      Not checked in the free run.

    • cites [48] D Centola, 2010, The spread of behavior in an online social network experiment, Science source

      Not checked in the free run.

38 of 38

  • The authors state this as an assumption.

    If two users share similar political ideologies, they can endorse and subsequently share (i.e. retweet) the content of each other.

    “We identified relevant interactions by assuming that if two users share similar political ideologies, they can endorse and subsequently share (i.e. retweet) the content of each other.”
1 of 1

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • It has been proposed that at population level, dynamics of multiple interacting contagions are indistinguishable from social reinforcement.

    “Recently, it has been suggested that this challenging phenomenon might be due to the fact that, at population level, the dynamics of multiple interacting contagions are indistinguishable from social reinforcement 38 .”

    From Assessing the risks of ‘infodemics’ in response to COVID-19 epidemics (2020) · cites this paper as [38]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Human-user total omits the aboutme dataset

    Minor

    The human counts in the table sum to 11,238, not 8,775. The reported total omits the 2,463 humans in the aboutme row, which also makes the grand total too low.

    Reported 8775 · recomputed 11238

  • Figure 1 panel (e) is assigned two different contents

    Minor

    The results text identifies panel (e) as the time series, while the figure caption identifies panel (e) as the geographic map.

2 of 2

Cites

Flagged first, then by citations

14 of the 66 references couldn’t be matched to a known paper.

Cited by

Sorted by citations

Related

Authors
Stella, M., Cristoforetti, M., & De Domenico, M.
Affiliations
Fondazione Bruno Kessler (IT)
ORCID
Venue
PLOS ONE
Publisher
Public Library of Science (PLoS)
Published
2019-05-16
Volume
14
Issue
5
Pages
e0214210
ISSN
1932-6203 (ISSN-L)
DOI
10.1371/journal.pone.0214210
PMC
PMC6521989
OpenAlex
W2789973986
Semantic Scholar
b4c0b98c98cdc11ab223e9d9ad2a2f8b4835f389
Cited by
51
References
66
Citations by year
2026: 1 · 2025: 4 · 2024: 8 · 2023: 8 · 2022: 7 · 2021: 12 · 2020: 7 · 2019: 3 · 2018: 1
FWCI
15.33
Type
journal article
Language
en
Open access
Licence
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • Institutional Research Information System (Università degli Studi di Trento) (submittedVersion, cc-by)
  • Padua Research Archive (University of Padova) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • arXiv (Cornell University)
Topics
  • Social Sciences › Communication › Social Media and Politics
  • Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
  • Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Complex Network Analysis Techniques
Keywords
augmented human, political bots, online social interaction, information cascades, collective attention, information flow, digital augmentation, internet influence
MeSH terms
Humans, Interpersonal Relations, Internet, Social Networking, Social Media

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review