Page 348 - Invited Paper Session (IPS) - Volume 1
P. 348
IPS162 Pedro C. et al.
3. Solutions
All this calls for skills of critical statistical literacy. Following Weiland (2016),
critical literacy is related to individuals being able to write both the word and
1
world—transforming their lived realities through the power of literacy . Based
1
on the intersection between Statistical literacy and Critical literacy, the author
introduces the key elements of critical statistical literacy in terms of reading and
writing information that are crucial to critical citizenship in today’s data centric
societies. Among these key elements, Weiland (2016) identifies the following:
“evaluating the source, collection and reporting of statistical information and
how they are influenced and shaped by the author’s social position and
sociopolitical and historical lens” and “Communicating one’s social location,
subjectivity, and political context to others and how it shapes one’s meaning
making of the world when reporting results of a statistical investigation.”
The problem of misinformation online has reached a proportion where
companies like Google, Facebook, and Twitter are forced to intervene.
Solutions are needed both for machine based actions and human based skills.
Facebook has implemented a system to give more information on the source
where an external URL is propagated in a post (Facebook, 2017), Twitter has
removed several accounts who are bots and who were spreading fake news
(Timberg & Dwoskin, 2018), and Google is tackling the problem by improving
media literacy (England, 2019). Several independent initiatives to develop tools
for detecting misinformation have occurred with the “Fake News Challenge”, a
competition to developed machine learning algorithms for detecting a stance
of a claim (Riedel, Augenstein, Spithourakis, & Riedel, 2017) and SemEval task
4 whose goal was to detect if a piece of news was hyperpartisan (Kiesel et al.,
2018).
The research community has also been active on the topic. Several
solutions have been proposed to tackle problems of misinformation online.
More specifically, research has tackled the analysis of false news (Shao et al.,
2018) and their propagation on the network (Vosoughi, Roy, & Aral, 2018), the
detection of bot and spam accounts (Benevenuto, Magno, Rodrigues, &
Almeida, 2010; Chu, Gianvecchio, Wang, & Jajodia, 2012) and other users that
are responsible for the spreading of unreliable content (Guimaraes, Figueira,
& Torgo, 2018). The targets of research also include the classification of
misinformation using machine and deep learning (Ruchansky, Seo, & Liu, 2017;
Tacchini, Ballarin, Della Vedova, Moret, & de Alfaro, 2017) and the
development of systems that allow fact-checking of claims recurring to
external knowledge databases (Ciampaglia et al., 2015; Shiralkar, Flammini,
Menczer, & Ciampaglia, 2017).
1 For a deeper understanding of the concept of statistical literacy, see, for example, Gal (2003)
and Watson and Callingham (2003).
337 | I S I W S C 2 0 1 9