ArticlePDF Available

Abstract and Figures

In this work we study, on a sample of 2.3 million individuals, how Facebook users consumed different information at the edge of political discussion and news during the last Italian electoral competition. Pages are categorized, according to their topics and the communities of interests they pertain to, in a) alternative information sources (diffusing topics that are neglected by science and main stream media); b) online political activism; and c) main stream media. We show that attention patterns are similar despite the different qualitative nature of the information, meaning that unsubstantiated claims (mainly conspiracy theories) reverberate for as long as other information. Finally, we categorize users according to their interaction patterns among the different topics and measure how a sample of this social ecosystem (1279 users) responded to the injection of 2788 false information posts. Our analysis reveals that users which are prominently interacting with alternative information sources (i.e. more exposed to unsubstantiated claims) are more prone to interact with false claims.
Content may be subject to copyright.
1
Collective attention in the age of (mis)information
Delia Mocanu1, Luca Rossi1, Qian Zhang1, M´arton Karsai1,2Walter Quattrociocchi1,3
1 Laboratory for the Modeling of Biological and Socio-technical Systems, Northeastern
University, Boston, MA 02115 USA
2 Laboratoire de l’Informatique du Parall´elisme, INRIA-UMR 5668, IXXI, ENS de Lyon
69364 Lyon, France
3 Laboratory of Computational Social Science, IMT Alti Studi Lucca, 55100 Lucca, Italy
Corresponding Author E-mail: walter.quattrociocchi@imtlucca.it
Abstract
In this work we study, on a sample of 2.3 million individuals, how Facebook users consumed different
information at the edge of political discussion and news during the last Italian electoral competition. Pages
are categorized, according to their topics and the communities of interests they pertain to, in a) alternative
information sources (diffusing topics that are neglected by science and main stream media); b) online
political activism; and c) main stream media. We show that attention patterns are similar despite the
different qualitative nature of the information, meaning that unsubstantiated claims (mainly conspiracy
theories) reverberate for as long as other information. Finally, we categorize users according to their
interaction patterns among the different topics and measure how a sample of this social ecosystem (1279
users) responded to the injection of 2788 false information posts. Our analysis reveals that users which
are prominently interacting with alternative information sources (i.e. more exposed to unsubstantiated
claims) are more prone to interact with false claims.
Introduction
The World Economic Forum, in its 2013 report [1], has listed the “massive digital misinformation” as
one of the main risks for the modern society. People perceptions, knowledge, beliefs, and opinions about
the world and its evolution get (in)formed and modulated through the information they can access,
most of which coming from newspapers, television [2], and, more recently, the Internet. The world
wide web, more specifically social networks and micro-blogging platforms, have changed the way we can
pursue intellectual growth or shape ideas. In particular, large social networks, with their user-provided
content, have been facilitating the study of how the economy of attention leads to specific patterns for
the emergence, production, and consumption of information [3–5].
Despite the enthusiastic rhetoric about the ways in which new technologies have burst the interest
in debating political or social relevant issues [6–11], the role of the socio-technical system in enforcing
informed debates and their effects on the public opinion still remain unclear. Indeed, the emergence
of knowledge from this process has been dubbed collective intelligence [12–16], although we have be-
come increasingly aware of the presence of unsubstantiated or untruthful rumors. False information is
particularly pervasive on social media, fostering sometimes a sort of collective credulity.
In this respect, conspiracists tend to explain significant social or political aspects as plots conceived by
powerful individuals or organizations [17]. As these kind of arguments can sometimes involve the rejection
of science, alternative explanations are invoked to replace the scientific evidence. For instance, people who
reject the link between HIV and AIDS generally believe that AIDS was created by the U.S. Government
to control the African American population [18, 19]. Since unsubstantiated claims are proliferating over
the Internet, what could happen if they were used as the basis for policy making?
A multitude of mechanisms animate the flow and acceptance of false rumors [20], which in turn create
false beliefs that are rarely corrected once adopted by an individual [21–24]. The process of acceptance of
a claim (whether documented or not) may be altered by normative social influence or by the coherence
arXiv:1403.3344v1 [cs.SI] 13 Mar 2014
2
with the individual system of beliefs [25, 26]. Nonetheless, several tools have been recently designed
to help users disambiguate misinformation and false news [27, 28]. On the other hand, basic questions
remain on how the quality of (mis)information affects the economy of attention processes, concerning,
for example, the virality of information, its lifespan and the consumption patterns.
A large body of literature addresses the study of social dynamics on socio-technical systems [29–34];
here we consider the relationship between information sources and online political debates, limiting our
investigation to the period preceding the Italian elections of 2013, and focusing our attention on the
Italian Facebook groups formed around political activism and alternative news sources.
We observed several interesting phenomena such as the proliferation of political pages and alternative
information sources with the aim to exploit the Internet peculiarities to organize and convey the public
discontent (with respect to the crisis and the decisions of the national government). Furthermore, we
noticed the emergence of very distinct groups, namely trolls, building Facebook pages as a parodistic
imitation of both alternative information sources and online political activism. Their activities range
from controversial comments and posting satirical content mimicking alternative news sources, to the
fabrication of purely fictitious statements, heavily unrealistic and sarcastic. Not rarely, these memes
became viral and were used as evidence in online debates from political activists [35]. Inspired by these
lively and controversial social dynamics, we addressed the quantitative analysis of the interlink between
information sources and political activism on the web. In particular, we want to understand the selection
criteria of users mostly exposed to unsubstantiated claims.
This paper is structured as follows. We will first introduce our methodology of categorizing the Face-
book pages, by taking into account their self-description as well as the type of content they promote.
We concentrate on alternative news sources, online political activism, and also on all the national main
stream news journals that we could find to have an active page on Facebook. In the following sections,
through thorough quantitative analysis, we show that the attention patterns when faced with various
contents are similar despite the different qualitative nature of the information, meaning that unsubstan-
tiated claims reverberate as long as other, more verified, information. Finally, we measure how the social
ecosystem responded to the perturbation of false information injected by trolls. We find that a domi-
nant fraction of the users interacting with the troll memes is the one composed of users preeminently
interacting with alternative information sources – and thus more exposed to unsubstantiated claims.
Surprisingly, consumers of alternative news, which are the users trying to avoid the main stream media
’mass-manipulation’, are the most responsive to the injection of false claims.
Methods
Case study and data collection
The debate around relevant social issues spreads and persists over the web, leading to the emergence of
unprecedented social phenomena such as the massive recruitment of people around common interests,
ideas or political visions. Disentangling the many factors behind the influence of information sources on
social perception is far from trivial. Specific knowledge about the cultural and social context (even if
online) in which they manifest is fundamental. Hence, inspired by the success of political movements
over the Internet, we start our investigation focusing on the social dynamics around pages of political
activism on the Italian Facebook during the 2013 electoral campaign. On one hand, political activists
conveyed the public discontent on the government and the economic conditions on a public arena; on
the other hand, as the main stream media are considered to be manipulated, alternative information
sources were free to disseminate news neglected by mainstream media or by science. In addition, we
notice the activity of an emerging group of users, namely trolls, producing caricatural versions of the
stories diffused by alternative information sources and political activism pages. As an outcome of this
period of observation, we compile a list of the most important and active Facebook pages of alternative
3
information sources and political movements.
The dataset is composed of 50 public pages for which we download all the posts (and their respective
users interactions) in a time span of six months (from Sept 1st, 2012 to Feb 28th, 2013). The entire
data collection process is performed exclusively with the Facebook Graph API [36], which is publicly
available and which can be used through one’s personal Facebook user account. The pages from which
we download data are public Facebook entities (can be accessed by virtually anyone). Most of the user
content contributing to such pages is also public unless the user’s privacy settings specify otherwise. The
exact breakdown of the data is presented in Table 1. We provide brief descriptions for each page in
Supporting Information.
The categorization of the pages is based on their different social functions together with the type of
information they disseminate. The first class includes all pages (that we could verify) of main stream
newspapers; the second category consists of alternative information sources - pages which disseminate
controversial information, most often lacking supporting evidence and sometimes contradictory of the
official news (e.g. conspiracy theories, link between vaccines and autism etc). The third category is
that of self-organized online political movements – with the role of gathering users to publicly convey
discontent against the current political and socio-economic situation (i.e. one major political party in
Italy has most of its activity online).
For all classes the focus of our analysis is on the interaction of users with the public posts – i.e, likes,
shares, and comments.
Total Mainstream News Alternative News Political Activism
Distinct users 2,368,555 786,952 1,072,873 1,287,481
Pages 50 8 26 16
Posts 193,255 51,500 92,566 49,189
Likes 23,077,647 4,334,852 7,990,225 10,752,570
Comments 4,395,363 1,719,409 935,527 1,740,427
Likes to Comments 4,731,447 1,710,241 1,146,275 1,874,931
Table 1. Breakdown of Facebook dataset. Mainstream News: all the national newspapers present
on Facebook. Alternative News: pages which disseminate controversial information, most often
lacking supporting evidence and sometimes contradictory of the official news.. Political Activism:
gathering users to publicly convey discontent against the current political and socio-economic situation.
Finally, we got access to 2788 post ids from a troll Page [37]. All of these posts are caricatural version
of political activism and alternative news stories, with the peculiarity to include always false information.
Despite the small dimension (7430 unique users, 18212 likes, 11337 comments and 9549 likes to comment)
the page was able to trigger several viral phenomena, one of which reached 100Kshares. We use troll
memes to measure how the social ecosystem under investigation is responding to the injection of false
information.
Results and Discussion
Attention patterns
We start our analysis providing an outline of users’ attention patterns with respect to different topics
coming from distinct sources - i.e, alternative news, main stream media and political activism. As a first
measure, we count the number of interactions (comments, likes, or likes to comments) by users and plot
4
the cumulative distribution function (CDF) of the users’ activity on the various page categories in Figure
1. CDF shows that user interactions with posts on all different types of pages does not present significant
differences. The similarity is also conserved after further grouping comments and likes separately (see in
Supporting Information Figure 5 and Figure 6).
Figure 1. Users Activity. Cumulative distribution function (CDF) of users’ activity, grouped by
page type. An action can be a like, comment, or like to comment. The distributions are nearly identical.
Here, the social response is not affected by the topic nor by the quality of the information. Posts
containing unsubstantiated claims, or about political activism, as well as regular news, cannot be distin-
guished through simple statistic signatures based on user engagement patterns. These different topics
reverberate at the same way in this ecosystem.
As the potential of memes to trigger discussions can be quantified through comments, in order to have
a more precise picture of the users’ attention, we zoom in to the level of posts. This level of resolution
is useful to understand the temporal evolution of posts and for how long the debate on a topic persists,
using the comments as a first-order approximation of the level of interest.
In Figure 2 we show, for each page type, the probability density function of the post interest lifetime.
This measure is computed as the temporal distance between the first and last comment of the post.
Collective debates persist similarly (see Supporting Information section for further details), independently
of whether the topic is the product of an official or unofficial source.
5
Figure 2. Post lifetime. Probability density function, grouped by page type, of the temporal
distance between the first and last comments of the post. Posts with qualitatively different topics
(alternative information, political activism, and main stream news) show a similar behavior.
Given the social context of these groups, what potential does a hurtful meme harness? In other words,
how significant is the concurrent presence of users between different pages and how strong is the overlap?
Starting from the null-hypothesis, that each user has neither affiliation nor preference, we investigate
the interaction dynamics of the different communities by quantifying the users that are present in both
spaces. The result in Table 2 hints that indeed a considerable number of users interact with pages of
different classes. The political discussion and alternative news get informed dominantly from each other
rather than from mainstream media, while users of the first two sets are almost equally represented within
the followers of mainstream newspapers.
Class A Class B Common users (AB) Ratio (AB/A) Ratio (AB/B)
Political movement Alternative news 360,054 28.0% 33.6%
Political movement Mainstream news 254,893 19.8% 32.4%
Mainstream news Alternative news 278,337 35.4% 25.9%
Table 2. Common users between classes of pages. Many users make active contributions on pages with
very different profiles. Several members of the political discussion are involved on both alternative news
pages and main stream newspapers pages in comparable ways.
In this portion of the Italian Facebook ecosystem untruthful rumors spread and trigger debate, rep-
resenting an important part of the information flow animating the political scenario.
Response to false information
Above results reveal that users consume unsubstantiated and main stream news in similar ways. In our
study, both are consumed by users of political activism pages. Continuing our investigation, we want to
6
understand if this information context might affect the users’ selection criteria. Therefore, we measure
the reaction of users to a set of 2788 false information injected by a troll page - i.e, a page promoting
caricatural version of alternative news and political activism stories.
In order to perform this analysis, we applied a classification strategy aimed at discriminating typical
users for each one of the three categories of pages. In particular, we were interested in distinguishing users
based on their behavior. Having access to the 6 months historical likes, comments, and likes to comments
on all posts within the timeframe (and within the privacy restrictions), we quantify the interaction of
each user with the posts in each class. As we did this, the following assumptions were in place:
The topic of the post is coherent with the theme of the page on which it was published.
A user is interested in the topic of the post if he/she likes the post. A comment, although it reflects
interest, is more ambiguous, and, therefore, is not considered to express a positive preference of the
topic.
We neither have access to nor try to guess the page subscription list of the users, regardless of their
privacy settings. Every step of the analysis involves only the active (participating) users on each
page.
According to these assumptions, we use solely the likes to the posts. For instance, if a user likes 10
different posts on one or multiple pages of the same political movement, but that user never liked posts
of any other topic, we will label that user to be associated with the political movement. Since it is not
always the case that there is a clear preference, we have to take into account the random sampling bias
error - since our data set represents indeed a sample of the users’ Facebook activity. Given the limitations
of the API, the only information we have about the user is how that user interacted with the posts we
have downloaded.
The labeling algorithm for each user is to calculate the 95% confidence interval of percentage of likes
of posts in each topic. Only if the confidence interval of the preferred topic does not overlap the other
two topics, we assign the user a label. Although the true affiliation of the individual behind the end user
can be a subjective matter, we believe that filtering out versatile users allows us to focus precisely on the
rare, and more interesting, cases of interaction between highly polarized users.
In Figure 3 we illustrate for each page type, the respective contributions brought by labeled (polar-
ized) users. It is important to note that this measure is not designed to describe the overall affiliation of
the members of the page. The fractions are computed by taking all the posts from a class and counting
percentage of users coming from each profile. Posts from alternative information sources and political ac-
tivism pages present a clear supremacy of the predominant class of users with, respectively, 45% and 49%
of the dominant class. Not surprisingly, mainstream media pages, present a more balanced distribution of
user classes, as their purpose is to communicate neutral information. However, users labeled as political
activists are more active on alternative information pages than on mainstream newspapers. In turn, users
labeled as mainstream media adepts are in minority on both alternative and activist pages. According to
this partitioning of the information space, now we are able to distinguish interactions occurring between
users pertaining to different regions of the ideological space.
7
Figure 3. For each page type: fractions of users with strong affiliations.
Given the outline of users distribution within the various classes, we want to see which users are more
responsive to the injection of false information in terms of interaction. As before, we cannot use the
comments as discriminators, as they can represent either positive or negative feedbacks with respect to
the published topic. Therefore, we focus only on the users liking 2788 troll posts.
As previously mentioned, troll posts are related to arguments debated by political activists or on
alternative information sources but with a clear parodistic flavor. For instance, one of the most popular
memes that explicitly spread a false rumor (in text form) reads: Italian Senate voted and accepted (257
in favor and 165 abstentions) a law proposed by Senator Cirenga aimed at funding with 134 billion Euros
the policy-makers to find a job in case of defeat in the political competition. We were able to easily verify
that this meme contains at least four false statements: the name of the senator, the total number of
votes is higher than possible, the amount of money (more than 10% of Italian GDP) as well as the law
itself. This meme was created by a troll page and, on the wave of public discontent against italian policy-
makers, quickly became viral, obtaining about 35,000 shares in less than one month. Shortly thereafter,
the image was downloaded and reposted (with the addition of a commentary) by a page describing itself
as being focused on political debate. Nowadays, this meme is among the arguments used by protesters
manifesting in several Italian cities. This is a striking example of the large scale effect of misinformation
diffusion on the opinion formation process. As shown in Figure 4 by counting the polarized users that
liked the posts, we find that the most susceptible users to interact with false information are those that
are mostly exposed and interacting with unsubstantiated claims (i.e. posts on alternative information
pages).
8
Figure 4. Social response to intentionally injected false information. Labels represent user affiliation.
The users more responsive to the injection of false information are the ones having strong affiliation
alternative information sources.
According to our results, users with strong preferences for alternative information sources, perhaps
motivated by the will to avoid the manipulation played by mainstream media controlled by the govern-
ment, are more susceptible to false information.
Conclusions
Conspiracists generally tend to explain a significant social or political aspect as a secret plot by powerful
individuals or organizations [17] and their activity is proliferating over the web. This study provides a
genuine outline of the online social dynamics and, in particular, on the effect of Facebook on bursting
the diffusion of false beliefs when truthful and untruthful rumors coexist.
In this work, we perform a case study aiming to understand the interlink between political discussion
and information on the web. The portion of Facebook we analyzed presents a complex set of social
interactions. Several cultures coexist, each one competing for the attention of users. Specifically, we
observe a strong interaction between political discussion and information sources (either alternative or
main stream). Most of the online activism Facebook pages contain claims that mainstream media is
manipulated by higher entities (and thus the information is be not neutral or reliable). Such an an-
tagonism makes any kind of persuasion process, even if based on more solid information, very difficult.
As a response to partisan debates, the emergent groups of trolls began to provide parodistic imitations
of a wide range of online partisan topics. Despite the evident parodistic (and sometimes paradoxical)
contents, not rarely, troll memes fomented animated debates and diffused through the community as any
other information would. Through statistical analysis, we find that the consumption patterns are similar
despite the different nature of the information. Finally, in order to uncover more characteristics of the
process, we distinguished users with strong affiliations and observed their respective interaction patterns,
as well as with false information inoculated in that portion of the Facebook ecosystem. We find that,
out of the 1279 labeled users interacting with the troll memes, a dominant percentage (56% , as opposed
to 26% and 18% for other groups) is constituted of users preeminently interacting with alternative infor-
mation sources and thus more exposed to unsubstantiated claims. The results of our study raise a real
warning, as the higher the number of circulating unsubstantiated claims is, the more users will be biased
in selecting contents.
9
Acknowledgments
Funding for this work was provided by the authors’ institutions (IMT Lucca Institute for Advanced
Studies, Northeastern University), EU FET project MULTIPLEX nr.317532. The funders had no role in
study design, data collection and analysis, decision to publish, or preparation of the manuscript.
We want to thank Alessandro Vespignani, Rosaria Conte, Mario Paolucci, Santo Fortunato, Brian
Keegan, Piotr Sapiezynski and Gianni Riotta for useful discussions
Special thanks go to Giulia Borrione, Dino Ballerini, Elio Gabalo, Monica D’Agruma, Stephanie
Ponzo, Giacomo Sorbi, Alpi Stefano, Umberto Mentana, Salvatore Previti for pointing out the phe-
nomenon of misinformation and for precious suggestions.
Authors Contribution
Conceived and designed the experiments: WQ DM. Performed the experiments: WQ. Analyzed the data:
WQ DM QZ. Contributed reagents/materials/analysis tools: DM LR QZ. Wrote the paper: WQ DM LR
QZ MK.
References
1. Howell L (2013) Digital wildfires in a hyperconnected world. In: Report 2013. World Economic
Forum.
2. Mccombs ME, Shaw DL (1972) The Agenda-Setting Function of Mass Media. The Public Opinion
Quarterly 36: 176–187.
3. Lanham RA (2007) The Economics of Attention: Style and Substance in the Age of Information.
University Of Chicago Press.
4. Qazvinian V, Rosengren E, Radev DR, Mei Q (2011) Rumor has it: Identifying misinformation
in microblogs. In: Proceedings of the Conference on Empirical Methods in Natural Language
Processing. Stroudsburg, PA, USA: Association for Computational Linguistics, EMNLP ’11, pp.
1589–1599. URL http://dl.acm.org/citation.cfm?id=2145432.2145602.
5. Dow PA, Adamic LA, Friggeri A (2013) The anatomy of large facebook cascades. In: Kiciman E,
Ellison NB, Hogan B, Resnick P, Soboroff I, editors, ICWSM. The AAAI Press.
6. Guillory J, Spiegel J, Drislane M, Weiss B, Donner W, et al. (2011) Upset now?: emotion contagion
in distributed groups. In: Proceedings of the SIGCHI Conference on Human Factors in Computing
Systems. New York, NY, USA: ACM, CHI ’11, pp. 745–748. doi:10.1145/1978942.1979049. URL
http://doi.acm.org/10.1145/1978942.1979049.
7. Bekkers V, Beunders H, Edwards A, Moody R (2011) New media, micromobilization, and political
agenda setting: Crossover effects in political mobilization and media usage. The Information
Society 27: 209–219.
8. Gonzalez-Bailon S, Borge-Holthoefer J, Rivero A, Moreno Y (2011) The dynamics of protest re-
cruitment through an online network. Scientific Report .
9. Garcia D, Mendez F, Serd¨ult U, Schweitzer F (2012) Political polarization and popularity in online
participatory media: an integrated approach. In: Proceedings of the first edition workshop on
Politics, elections and data. New York, NY, USA: ACM, PLEAD ’12, pp. 3–10. doi:10.1145/
2389661.2389665. URL http://doi.acm.org/10.1145/2389661.2389665.
10
10. Crespi I (1997) The public opinion process. How the people speak. Lawrence Erlbaum Associates.
11. Lippmann W (1946) Public opinion. Penguin Books.
12. Levy P (1999) Collective Intelligence: Mankind’s Emerging World in Cyberspace. Perseus Pub-
lishing.
13. Buckingham Shum S, Aberer K, Schmidt A, Bishop S, Lukowicz P, et al. (2012) Towards a global
participatory platform. The European Physical Journal Special Topics 214: 109-152.
14. Levy P (2000) Collective Intelligence: Mankind’s Emerging World in Cyberspace.
15. Malone TW, Klein M (2007) Harnessing collective intelligence to address global climate change.
Innovations: Technology, Governance, Globalization 2: 15-26.
16. Shadbolt N, Hall W, Hendler JA, Dutton WH (2013) Web science: a new frontier. Physical and
Engineering Sciences 371.
17. Sunstein CR, Vermeule A (2009) Conspiracy theories: Causes and cures*. Journal of Political
Philosophy 17: 202–227.
18. Bogart LM, Thorburn S (2005) Are HIV/AIDS conspiracy beliefs a barrier to HIV prevention
among african americans? J Acquir Immune Defic Syndr 38: 213–8.
19. Kalichman SC (2009) Denying aids: Conspiracy theories, pseudoscience, and human tragedy : 228.
20. Kuklinski JH, Quirk PJ, Jerit J, Schwieder D, Rich RF (2000) Misinformation and the currency
of democratic citizenship. The Journal of Politics 62: 790–816.
21. Garrett RK, Weeks BE (2013) The promise and peril of real-time corrections to political mis-
perceptions. In: Proceedings of the 2013 conference on Computer supported cooperative work.
New York, NY, USA: ACM, CSCW ’13, pp. 1047–1058. doi:10.1145/2441776.2441895. URL
http://doi.acm.org/10.1145/2441776.2441895.
22. Meade M, Roediger H (2002) Explorations in the social contagion of memory. Memory & Cognition
30: 995-1009.
23. Koriat A, Goldsmith M, Pansky A (2000) Toward a psychology of memory accuracy. Annu Rev
Psychol 51: 481–537.
24. Ayers M, Reder L (1998) A theoretical review of the misinformation effect: Predictions from an
activation-based memory model. Psychonomic Bulletin & Review 5: 1–21.
25. Zhu B, Chen C, Loftus EF, Lin C, He Q, et al. (2010) Individual differences in false memory
from misinformation: Personality characteristics and their interactions with cognitive abilities.
Personality and Individual Differences 48: 889 - 894.
26. Frenda SJ, Nichols RM, Loftus EF (2011) Current Issues and Advances in Misinformation Research.
Current Directions in Psychological Science 20: 20–23.
27. Ennals R, Trushkowsky B, Agosta JM (2010) Highlighting disputed claims on the web. In: Proceed-
ings of the 19th international conference on World wide web. New York, NY, USA: ACM, WWW
’10, pp. 341–350. doi:10.1145/1772690.1772726. URL http://doi.acm.org/10.1145/1772690.
1772726.
11
28. McKelvey K, Menczer F (2013) Truthy: Enabling the study of online social networks. In: Proc.
CSCW ’13. URL http://arxiv.org/abs/1212.4565.
29. Onnela JP, Reed-Tsochas F (2010) Spontaneous emergence of social influence in online systems.
Proceedings of the National Academy of Sciences 107: 18375–18380.
30. Ugander J, Backstrom L, Marlow C, Kleinberg J (2012) Structural diversity in social contagion.
Proceedings of the National Academy of Sciences .
31. Lewis K, Gonzalez M, Kaufman J (2012) Social selection and peer influence in an online social
network. Proceedings of the National Academy of Sciences 109: 68–72.
32. Mocanu D, Baronchelli A, Gon¸calves B, Perra N, Zhang Q, et al. (2013) The twitter of babel:
Mapping world languages through microblogging platforms. PLOS ONE 8: e61981.
33. Adamic L, Glance N (2005) The political blogosphere and the 2004 u.s. election: Divided they
blog. In: In LinkKDD 05: Proceedings of the 3rd international workshop on Link discovery. pp.
36–43.
34. Kleinberg J (2013) Analysis of large-scale social and information networks. Philosophical Transac-
tions of the Royal Society A: Mathematical, Physical and Engineering Sciences 371.
35. Ambrosetti G (2013). I forconi: il senato ha approvato una legge per i par-
lamentari in crisi. chi non verr rieletto, oltre alla buonuscita, si beccher al-
tri soldi. sar vero? Website. URL http://www.linksicilia.it/2013/08/
i-forconi-il-senato-ha-approvato-una-legge-per-i-parlamentari-in-crisi-chi-non-verra-rieletto-oltre-alla-buonuscita-si-becchera-altri-soldi-sara-vero/.
Last checked: 19.01.2014.
36. Facebook (2013). Using the graph api. Website. URL https://developers.facebook.com/docs/
graph-api/using-graph-api/. Last checked: 19.01.2014.
37. (2012). Semplicementeme. Website. URL hhttps://www.facebook.com/pages/
Semplicemente-me/286003974834081?fref=ts. Last checked: 19.01.2014.
12
1 Supporting Information
Data Collection
The entire data collection process is performed exclusively with the Facebook Graph API, which is
publicly available and which can be used through one’s personal Facebook user account.
The pages from which we download data are public Facebook entities (can be accessed by virtually
anyone). Most of the user content contributing to such pages is also public unless the user’s privacy
settings specify otherwise.
Using the Graph API, one is able to download the feed of a page, along with the content that would
otherwise be visible though regular browsing. There are, however, several limitations to this process
which we will attempt to describe in detail later in this section.
Our approach: for each page we access the feed between Sept 1st, 2012 and Feb 28th, 2013; that is, we
download all the public posts created in this period and the related content. Specifically, we can identify
the users that liked the post, we can access the comments (content, user, time). Other fields such as post
author, post creation time, and post type (e.g. photo, link, video, plain text) are also available.
Note 1: When a post points to a photo, it has two possible forms. The first is when the photo upload
is the post itself. The second type is when the post actually links to a photo that already existed as a
static object (i.e. it pertains to an album and has a fixed address as well as a unique identifier). In the
latter case, if one were to click on the post, one is redirected to the unique address of the photo in case.
Therefore, it is possible to have different posts pointing to the same fixed object (photo or video). It is
important at this point to make the distinction between post and object. A post can accrue comments,
likes, or shares that are separate from the comments, likes, and shares of the object it points to. For this
reason, when we encounter a post that also points to an object, we download the data associated with
the latter as well. Naturally, an object has the benefit of having accrued more interest over time.
The rate limits that we encounter with the Graph API restrict us to only being able to access the last
5000 shares of an object or post. The likes and comments, however, are not subject to such rate limits.
At all times the privacy limits are in effect, so the data we obtain is usually smaller than the total
number theoretically available. For instance, it is possible that we only get about 4000 of the last shares
of an object. Similarly, we can only get partial data on likes and comments. Since we do have access to
the total number of shares, likes, comments, according to our observation, about 20% of the user actions
are invisible to us. This property is specific to our users/set of pages and does not necessarily accurately
reflect the properties of the rest of the Facebook community outside our dataset.
Note 2: Within the privacy restrictions, one is able to access the branches of each share action, should
they exist. The reverse process (upstream) is not possible using the Graph API, unlike the case of direct
web surfing. For this reason, along with the privacy settings of some users, it is virtually impossible to
reconstruct the complete sharing tree of a popular object (with more than 5000 shares).
Note 3: One important limitation of the Graph API, which manual surfing does not present, is that
one cannot access any user profile information, even if such information is otherwise public. Consequently,
we only have access to the unique id and to the name of each user - but not to their location, for instance.
13
Attention patterns: likes and comments distributions
Figure 5. Users’ likes cumulative distribution function, grouped by page type.
Figure 6. Users’ comments cumulative distribution function, grouped by page type.
14
Post lifetime in the different classes
In this section we provide further details about the post lifetime in the various classes. We consider
lifetime as the distance in time between the first and the last comment to a post. Here we provide a
comparisons among the distributions of the lifetime for each post according to the various classes. In
Figures 7, 8 and 9 we show the quantile-quantile plot of the normalized distributions.
Figure 7. Quantile vs Quantlie of post lifetime in alternative news and mainstream news.
Figure 8. Quantile vs Quantlie of post lifetime in alternative news and political movements.
15
Figure 9. Quantile vs Quantlie of post lifetime in mainstream news and political movements.
In Table 4 we report the results of Kolmogorov-Smirnov test. Notice that since the KS test loses
precision in heavy tailed data we perform the analysis by cutting the tail at 0.1 on the normalized
distributions.
Class A Class B p-value
Political movement Alternative news 0.9253
Political movement Mainstream news 0.1087
Mainstream news Alternative news 0.156
Table 3. Kolmogorov-Smirnov test on the different distributions coming from the different classes.
In Table 4 we report the results of t test. As in the KS case we cut the tail at 0.1 of the normalized
distributions.
Class A Class B p-value
Political movement Alternative news 0.3249
Political movement Mainstream news 0.9863
Mainstream news Alternative news 0.2768
Table 4. T-test on the different distributions coming from the different classes.
16
List of pages
Category Page Followers Description
Alt.Inf Libert di Stampa 289k Supporting the right of expression.
Reporting news mostly oriented on
politics (lobbies and government).
Alt. Inf Lo Sai Economia 7k Delivering news about economics
(lobbies, signorage, free masons and
so forth)
Alt. Inf Informazione Libera 1m Pointing out all the corruption of the
political “lobby” and the current dra-
matic situation of the middle class
Alt. Inf Il Radar 15k News Magazine supporting the Right
party. Pointing out all the paradox
(sometimes exagerated) of the oppo-
site parties
Alt. Inf NeoVitruvian 5k NWO order, the miracle of the free
energy by Nicola Tesla, Illuminati etc
etc
Alt. Inf NoCensura.com 480k All the things that are “intentionally
ignored by manipulated media” (sig-
norage, opinion manipulation..)
Alt. Inf Lo Sai Salute e Alimen-
tazione
18k Information and discussion against
the traditional medicine practices
(Vaccines and Autism, OGM, etc)
Alt. Inf Terra Real Time 38k Signorage, Alternative healtcare,
David Icke thesis, Chemical Trails...
Alt. Inf Lo Sai Chemtrails 6k The cronichle of the chem trails....
Alt. Inf Signoraggio Bancario 6k All news and info about the advances
of the NWO plans and the world dom-
inated by the banks (main thesis BCE
and Federal Reserve are private com-
panies creating the public debt with
an escamotage)
Alt. Inf Nuovo Ordine Mondiale 3k Discussion about the NWO plans and
its evidences
Alt.Inf L’altra Notizia 70k All the information neglected by the
manipulated media
Alt. Inf Contro Copertina 10k Mainly populistic arguments to con-
vey public discontent by the wave of
indignation
Alt. Inf Haarp Controllo Globale 3k Diffusion of information about the ex-
istence of secret plans to destabilize
the world by causing earthquakes and
by poisoning humanity with barium
and other not well specified substan-
cies released by airplanes.
17
Alt. Inf Stop alle Scie Chimiche 14k Pointing out the Chemical Trails
plans
Alt. Inf LoSai.Eu 135k One of the most active pages in dis-
seminating all information neglected
by main stream media
Alt.Inf Verit 11 Settembre 14k Supporting alternative thesis about
9/11 official version
Alt.Inf Uniti contro le multi-
nazionali
17k Major corporation are poisoning the
world and natural medications has to
replace medicines (curing cancer with
bicarbonatum)
Alt.Inf Condividi la Conoscenza 95k News to about all the facts (mainly
politics) neglect by main stream me-
dia
Alt. Inf Informare per resistere 762k One of the most active page to dif-
fuse information neglected by the me-
dia through the web
Alt. Inf Contro Informazione Al-
ternativa
45k News with a populistic tone in par-
ticular against the political actions of
the government
Alt.Inf Orwell 2012 2k NWO plans and opinion manipulation
Alt.Inf HAARP 2k Diffusion of information about the ex-
istence of secret plans to destabilize
the world by causing earthquakes and
by poisoning humanity with barium
and other not well specified substan-
cies released by airplanes.
Alt. Inf Contro l’informazione ma-
nipolata
103k All the information neglected by the
manipulated media controlled by the
governments and lobbies
Alt. Inf Informare Contro Infor-
mando
32k News to about all the facts (mainly
politics) neglect by main stream me-
dia
Alt.Inf. Smascheriamo gli illumi-
nati
3k Pointing out all the symbols and
subliminal messages from the main
stream media delivered by the illumi-
nati
Troll Semplicemente me 5k Diffusing memes as a paradoxal imi-
tation of alternative information and
political movements
Main Stream Media Il Giornale 125k National official news paper near a the
center right party (PDL) of Berlus-
coni
Main Stream Media La Repubblica 1.3m National news paper close to the cen-
ter left party (PD) and the most dif-
fused italian journal
Main Stream Media Il Fatto Quotidiano 1.1m National news paper near at the Five
Star Movement (M5S) of Beppe Grillo
18
Main Stream Media Il Manifesto 69k National news paper near to the left
party
Main Stream Media Il Corriere della Sera 1m National News Paper
Main Stream Media La Stampa 131k National News Paper
Main Stream Media Il Sole 24 ore 267k National News Paper more oriented
on the economic and financial systems
Main Stream Media Il Messaggero 183k National News Paper
Political Activism Incazzati contro la casta 78k Convey the public discontent against
the socio-economic situation
Political Activism RNA-Rete Anti Nucleare 47k Inform and sensibilization against the
Nuclear Energy in Italy
Political Activism Indignados Italia 44k Discussion about the current socio-
economic situation
Political Activism Questa l’Italia 118k Convey the public discontent against
the socio-economic situation
Political Activism NewApocalypse Not On-
line
Convey the debate against Signorage
and the NWO
Political Activism Qelsi 90k Against the left parties
Political Activism Partigiani del III Millen-
nio
Not more
online
Convey the public discontent against
the socio-economic situation
Political Activism Vogliamo i Parlamentari
in Carcere...
45k Convey the public discontent against
the socio-economic situation (arrest-
ing all policy makers)
Political Activism Adesso Fuori dai Coglioni 560k Convey the public discontent against
the socio-economic situation
Political Activism Forza Nuova 50k Convey the public discontent against
the socio-economic situation (extreme
right party)
Political Activism Vota Casa Pound 48k Convey the public discontent against
the socio-economic situation (extreme
right party)
Political Activism Gruppo Free Italy 15k Convey the public discontent against
the socio-economic situation
Political Activism Catena Umana Attorno al
Parlamento
105k Convey the public discontent against
the socio-economic situation
Political Activism Non chiamateli Politici ma
criminali
45k Convey the public discontent against
the socio-economic situation
Political Activism BeppeGrillo 1m The page of the M5S Leader propos-
ing the free circualtion of the infor-
mation on the Internet as a major
revolution. A good example of e-
participation.
Political Activism Alice nel paese delle mer-
daviglie
70k Convey the public discontent against
the socio-economic situation
... Numerous researches have focused on reviewing the characteristics of rumor propagation, analyze features (Friggeri et al. [4]). One study discloses that people exposed to other info sources than the official ones are more susceptible to untruth claims (Mocanu et al. [11]). The study has also been carried out to identify distrustful memes (Ratkiewicz et al. [18]) while spread through Digg and Twitter platforms (Lerman and Ghosh [8]). ...
... Proof. Assume that equation (11) holds. Then ...
... Due to social media's easy availability and convenience, information spreads more rapidly and widely through these platforms than its traditional counterparts (9,10). Moreover, the resulting mass of user-provided content promotes vast recruitment of people around shared interests, worldviews, and narratives, thus influencing the evolution of public opinion (11) and further enabling rumors to thrive. In 2013, the World Economic Forum described web-based rumors as 'digital wildfire' and accentuated the risks they pose to modern society (12). ...
Article
Full-text available
Background In the early stage of the COVID-19 outbreak in China, several social rumors in the form of false news, conspiracy theories, and magical cures had ever been shared and spread among the general public at an alarming rate, causing public panic and increasing the complexity and difficulty of social management. Therefore, this study aims to reveal the characteristics and the driving factors of the social rumors during the COVID-19 pandemic.Methods Based on a sample of 1,537 rumors collected from Sina Weibo's debunking account, this paper first divided the sample into four categories and calculated the risk level of all kinds of rumors. Then, time evolution analysis and correlation analysis were adopted to study the time evolution characteristics and the spatial and temporal correlation characteristics of the rumors, and the four stages of development were also divided according to the number of rumors. Besides, to extract the key driving factors from 15 rumor-driving factors, the social network analysis method was used to investigate the driver-driver 1-mode network characteristics, the generation driver-rumor 2-mode network characteristics, and the spreading driver-rumor 2-mode characteristics.ResultsResearch findings showed that the number of rumors related to COVID-19 were gradually decreased as the outbreak was brought under control, which proved the importance of epidemic prevention and control to maintain social stability. Combining the number and risk perception levels of the four types of rumors, it could be concluded that the Creating Panic-type rumors were the most harmful to society. The results of rumor drivers indicated that panic psychology and the lag in releasing government information played an essential role in driving the generation and spread of rumors. The public's low scientific literacy and difficulty in discerning highly confusing rumors encouraged them to participate in spreading rumors.Conclusion The study revealed the mechanism of rumors. In addition, studies involving rumors on different emergencies and social platforms are warranted to enrich the findings.
... For example, [161] show that author information can be a useful feature for fake news detection, and [69] attempt to determine the veracity of a claim based on the conversation it sparks on Twitter as one of the RumourEval tasks. The Facebook analysis of [173] shows that unsubstantiated claims spread as widely as well-established ones, and that user groups predisposed to conspiracy theories are more open to sharing the former. Similarly, [1], [154], [164], and [237] model the spread of (mis-)information, while [41] and [178] propose algorithms to limit its spread. ...
Thesis
Full-text available
With the growing importance of the World Wide Web, the major challenges our society faces are also increasingly affecting the digital areas of our lives. Some of the associated problems can be addressed by computer science, and some of these specifically by data-driven research. To do so, however, requires to solve open issues related to archive quality and the large volume and variety of the data contained. This dissertation contributes data, algorithms, and concepts towards leveraging the big data and temporal provenance capabilities of web archives to tackle societal challenges. We selected three such challenges that highlight the central issues of archive quality, data volume, and data variety, respectively: (1) For the preservation of digital culture, this thesis investigates and improves the automatic quality assurance of the web page archiving process, as well as the further processing of the resulting archive data for automatic analysis. (2) For the critical assessment of information, this thesis examines large datasets of Wikipedia and news articles and presents new methods for automatically determining quality and bias. (3) For digital security and privacy, this thesis exploits the variety of content on the web to quantify the security of mnemonic passwords and analyzes the privacy-aware re-finding of the various seen content through private web archives.
... The main intuition is that misinformation creators have a different writing style seeking to maximize reading, sharing and, in general, maximizing virality. This is important because views and engagement in social networks are closely related to virality, and being repeatedly exposed to misinformation increases the likelihood of believing in false claims (Bessi et al., 2015;Mocanu et al., 2015). ...
Article
Full-text available
Not all misinformation is created equal. It can adopt many different forms like conspiracy theories, fake news, junk science, or rumors among others. However, most of the existing research does not account for these differences. This paper explores the characteristics of misinformation content compared to factual news—the “fingerprints of misinformation”—using 92,112 news articles classified into several categories: clickbait, conspiracy theories, fake news, hate speech, junk science, and rumors. These misinformation categories are compared with factual news measuring the cognitive effort needed to process the content (grammar and lexical complexity) and its emotional evocation (sentiment analysis and appeal to morality). The results show that misinformation, on average, is easier to process in terms of cognitive effort (3% easier to read and 15% less lexically diverse) and more emotional (10 times more relying on negative sentiment and 37% more appealing to morality). This paper is a call for more fine-grained research since these results indicate that we should not treat all misinformation equally since there are significant differences among misinformation categories that are not considered in previous studies.
... It is also theorized that the individual assessment of one's own risk can be influenced by the information received and the social context in which the individual lives and interacts [29]. It is well-established that many people seek information that supports their convictions-a phenomenon called "confirmation bias"-and that misinformation is frequently shared in echo chambers of like-minded groups of individuals [30][31][32]. As such, within these groups, the social expectation of not getting the vaccine may play a critical role in shaping vaccine intentions. ...
Article
Full-text available
The COVID-19 pandemic has highlighted the adverse consequences created by an infodemic, specifically bringing attention to compliance with public health guidance and vaccine uptake. COVID-19 vaccine hesitancy is a complex construct that is related to health beliefs, misinformation exposure, and perceptions of governmental institutions. This study draws on theoretical models and current data on the COVID-19 infodemic to explore the association between the perceived risk of COVID-19, level of misinformation endorsement, and opinions about the government response on vaccine uptake. We surveyed a sample of 2697 respondents from the US, Canada, and Italy using a mobile platform between 21–28 May 2021. Using multivariate regression, we found that country of residence, risk perception of contracting and spreading COVID-19, perception of government response and transparency, and misinformation endorsement were associated with the odds of vaccine hesitancy. Higher perceived risk was associated with lower odds of hesitancy, while lower perceptions of government response and higher misinformation endorsement were associated with higher hesitancy.
... It is also theorized that the individual assessment of one's own risk can be influenced by the information received, and the social context in which the individual lives and interacts [27]. It is well-established that many people seek information that supports their convictions -a phenomenon called "confirmation bias" [28,29]. Misinformation is frequently shared in echo chambers of like-minded groups of individuals [30]. ...
Preprint
Full-text available
The COVID-19 pandemic has highlighted the adverse consequences created by an infodemic specifically on compliance with public health guidance and vaccine uptake. COVID-19 vaccine hesitancy is a complex construct that is related to health beliefs, misinformation exposure, and perceptions of governmental institutions. This study draws on theoretical models and current data on the COVID-19 infodemic to explore the association between perceived risk of COVID-19, levels of misinformation endorsement, and opinions about the government response on vaccine uptake. We surveyed a sample of 2,697 respondents from the US, Canada, and Italy using a mobile platform between 21-28 May, 2021. Using multivariate regression, we found that country of residence, risk perception of contracting and spreading COVID-19, perception of government response and transparency, and misinformation endorsement was associated with the odds of vaccine hesitancy. Higher perceived risk was associated with lower odds of hesitancy, while lower perceptions of government response, and higher misinformation endorsement were associated with higher hesitancy.
... While it is unclear which NFO group would be more susceptible to misinformation in general and be misinformed about the coronavirus in particular, Lee and Shin (2021) argued that NFO would affect people's vulnerability to misinformation through media consumption. Given that those who consume alternative, partisan media are more likely to encounter misinformation (Mocanu et al., 2015) and believe misinformation that favors the political party they support (Hutchens et al., 2021), the moderate-active NFO group, which often seeks out partisan information (Camaj, 2014(Camaj, , 2019, is more likely to be influenced by misinformation. Further agenda-setting studies suggest a strong connection between partisan media and misinformation, with partisan media both increasing the viral spread of misinformation (Rojecki & Meraz, 2016) and setting and reacting to the agenda of fake news sites in the 2016 election (Guo & Vargo, 2020;Vargo et al., 2018). ...
Article
The present study explores the relationship between the need for orientation (NFO) and knowledge/misperception about COVID-19 using a two-wave panel survey of U.S. adults (W1: N = 1,119; W2: N = 543). The findings suggest that moderate-active NFO rather than high NFO better predicts individuals’ level of knowledge and misperception. We also found that different media use (vertical media and horizontal media) and individuals’ epistemic beliefs (intuitionism and rationalism) have distinct implications for knowledge and misperception about COVID-19.
Preprint
Full-text available
This is an indexed collection and depiction of typologies and taxonomies of misinformation, disinformation, and fake news.
Conference Paper
Full-text available
Computer scientists have responded to the high prevalence of inaccurate political information online by creating systems that identify and flag false claims. Warning users of inaccurate information as it is displayed has obvious appeal, but it also poses risk. Compared to post-exposure corrections, real-time corrections may cause users to be more resistant to factual information. This paper presents an experiment comparing the effects of real-time corrections to corrections that are presented after a short distractor task. Although real-time corrections are modestly more effective than delayed corrections overall, closer inspection reveals that this is only true among individuals predisposed to reject the false claim. In contrast, individuals whose attitudes are supported by the inaccurate information distrust the source more when corrections are presented in real time, yielding beliefs comparable to those never exposed to a correction. We find no evidence of real-time corrections encouraging counterargument. Strategies for reducing these biases are discussed.
Article
Full-text available
The FuturICT project seeks to use the power of big data, analytic models grounded in complexity science, and the collective intelligence they yield for societal benefit. Accordingly, this paper argues that these new tools should not remain the preserve of restricted government, scientific or corporate élites, but be opened up for societal engagement and critique. To democratise such assets as a public good, requires a sustainable ecosystem enabling different kinds of stakeholder in society, including but not limited to, citizens and advocacy groups, school and university students, policy analysts, scientists, software developers, journalists and politicians. Our working name for envisioning a sociotechnical infrastructure capable of engaging such a wide constituency is the Global Participatory Platform (GPP). We consider what it means to develop a GPP at the different levels of data, models and deliberation, motivating a framework for different stakeholders to find their ecological niches at different levels within the system, serving the functions of (i) sensing the environment in order to pool data, (ii) mining the resulting data for patterns in order to model the past/present/future, and (iii) sharing and contesting possible interpretations of what those models might mean, and in a policy context, possible decisions. A research objective is also to apply the concepts and tools of complexity science and social science to the project’s own work. We therefore conceive the global participatory platform as a resilient, epistemic ecosystem, whose design will make it capable of self-organization and adaptation to a dynamic environment, and whose structure and contributions are themselves networks of stakeholders, challenges, issues, ideas and arguments whose structure and dynamics can be modelled and analysed. Graphical abstract
Conference Paper
Full-text available
When users post photos on Facebook, they have the option of allowing their friends, followers, or anyone at all to subsequently reshare the photo. A portion of the billions of photos posted to Facebook generates cascades of reshares, enabling many additional users to see, like, comment, and reshare the photos. In this paper we present characteristics of such cascades in aggregate, finding that a small fraction of photos account for a significant proportion of reshare activity and generate cascades of non-trivial size and depth. We also show that the true influence chains in such cascades can be much deeper than what is visible through direct attribution. To illuminate how large cascades can form, we study the diffusion trees of two widely distributed photos: one posted on President Barack Obama's page following his reelection victory, and another posted by an individual Facebook user hoping to garner enough likes for a cause. We show that the two cascades, despite achieving comparable total sizes, are markedly different in their time evolution, reshare depth distribution, predictability of subcascade sizes, and the demographics of users who propagate them. The findings suggest not only that cascades can achieve considerable size but that they can do so in distinct ways. Copyright © 2013, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
Article
Full-text available
Conspiracy theories (CTs) can take many forms and vary widely in popularity, the intensity with which they are believed and their effects on individual and collective behavior. An integrated account of CTs thus needs to explain how they come to appeal to potential believers, how they spread from one person to the next via communication, and how they motivate collective action. We summarize these aspects under the labels of stick, spread, and action. We propose the quasi-religious hypothesis for CTs: drawing on cognitive science of religion, social representations theory, and frame theory. We use cognitive science of religion to describe the main features of the content of CTs that explain how they come to stick: CTs are quasi-religious representations in that their contents, forms and functions parallel those found in beliefs of institutionalized religions. However, CTs are quasi-religious in that CTs and the communities that support them, lack many of the institutional features of organized religions. We use social representations theory to explain how CTs spread as devices for making sense of sudden events that threaten existing worldviews. CTs allow laypersons to interpret such events by relating them to common sense, thereby defusing some of the anxiety that those events generate. We use frame theory to explain how some, but not all CTs mobilize collective counter-conspiratorial action by identifying a target and by proposing credible and concrete rationales for action. We specify our integrated account in 13 propositions.
Conference Paper
Full-text available
We present our approach to online popularity and its applications to political science, aiming at the creation of agentbased models that reproduce patterns of popularity in participatory media. We illustrate our approach analyzing a dataset from Youtube, composed of the view statistics and comments for the videos of the U.S. presidential campaigns of 2008 and 2012. Using sentiment analysis, we quantify the collective emotions expressed by the viewers, finding that democrat campaigns elicited more positive collective emotions than republican campaigns. Techniques from computational social science allow us to measure virality of the videos of each campaign, to find that democrat videos are shared faster but republican ones are remembered longer inside the community. Last we present our work in progress in voting advice applications, and our results analyzing the data from choose4greece.com. We show how we assess the policy differences between parties and their voters, and how voting advice applications can be extended to test our agentbased models.
Article
Full-text available
Large scale analysis and statistics of socio-technical systems that just a few short years ago would have required the use of consistent economic and human resources can nowadays be conveniently performed by mining the enormous amount of digital data produced by human activities. Although a characterization of several aspects of our societies is emerging from the data revolution, a number of questions concerning the reliability and the biases inherent to the big data "proxies" of social life are still open. Here, we survey worldwide linguistic indicators and trends through the analysis of a large-scale dataset of microblogging posts. We show that available data allow for the study of language geography at scales ranging from country-level aggregation to specific city neighborhoods. The high resolution and coverage of the data allows us to investigate different indicators such as the linguistic homogeneity of different countries, the touristic seasonal patterns within countries and the geographical distribution of different languages in multilingual regions. This work highlights the potential of geolocalized studies of open data sources to improve current analysis and develop indicators for major social phenomena in specific communities.
Book
Extremism and thePsychologyof Uncertainty showcases cutting-edge scientific research on the extent to which uncertainty may lead to extremism. Contributions come from leading international scholars who focus on a wide variety of forms, facets and manifestations of extremist behavior. Systematically integrates and explores the growing diversity of social psychological perspectives on the uncertainty extremism relationship .Showcases contemporary cutting edge scientific research from leading international scholars. Offers a broad perspective on extremism and focuses on a wide variety of different forms, facets and manifestations. Accessible to social and behavioral scientists, policy makers and those with a genuine interest in understanding the psychology of extremism.
Book
No event of any significance in the world today – be it an unexpected election result, a terrorist attack, the death of a public figure, a meteorological anomaly, or the flu pandemic – takes place without generating at least a flutter of conspiracy speculations. Conspiracy Theories: A Critical Introduction offers a well informed, highly accessible, and thoroughly engaging introduction to conspiracy theories, discussing their nature and history, causes and consequences. Through a series of specific questions that cut to the core of conspiracism as a global social and cultural phenomenon, the book deconstructs the logic and rhetoric of conspiracy theories and analyses the broader social and psychological factors that contribute to their persistence in modern society. • What are the defining characteristics of conspiracy theories and how do they differ from legitimate inquiries into actual conspiracies? • How long have conspiracy theories been around and to what extent are contemporary versions similar to those of yesteryear? • Why do conspiracy theories all sound alike and what ensures their persistence in modern society? • What psychological benefits do conspiracy theories bring to those who subscribe to them? • Why are conspiracy theories so often mobilized by political forces whose agenda is antithetical to democratic politics?