Approximation to the map on disinformation and verification research in Spain: state of affairs

1. Introduction

Disinformation can cause significant damage to the lives of people and organizations. It can also impact health, finances and certainly has the potential to be detrimental to democracy. The crisis that emerged from the COVID-19 pandemic has fueled the debate on the public’s need to protect ourselves against the spread of hoaxes and false news, which during times of tension and uncertainty can seriously affect social stability and peaceful coexistence. This threat is also capable of shaking the foundations on which Western democracies stand, as this depends upon the trust that citizens place in their representatives and public institutions.

At the same time, the democratization of information has caused a proliferation of so-called fake news. Indeed, "as access to information has grown exponentially, so has access to disinformation" (Ortiz-de-Guinea-Ayala and Martín-Sáez, 2019).

In recent years, research and publications on the concept of fake news have increased considerably; (Blanco et al., 2019), including studies carried out in the Ibero-American sphere (Guallar et al., 2020). This has become even more apparent in the face of Covid-19 and the so-called global infodemic (García-Marín, 2020).

Factors such as polarization and digital misinformation (Jungherr and Schroeder, 2021) led to greater concern about the manipulation of citizens, especially during elections through strategies such as the creation of an artificial public opinion that could encourage conflict in politics (Frost, 2020).

In this sense, the work of verification journalism is fundamental in providing accountability to society. It is also linked to trust in the State and the Public Administration power, as well as to human rights (Rodríguez Pérez, 2020). Manipulation in the dissemination of information, the use of biased data and arguments, including disinformation, the propagation of lies, half-truths and false news, are not new and have many precedents in the history of journalism (Jay, 2010).

Countries, institutions, governments and parties are currently dealing with post-truth. This term was coined by the sociologist Ralph Keyes (2004) to refer to the process of creative manipulation of the truth, which is embellished and reconfigured to suit audiences. In post-truth, facts are less important than the sensations and emotions they produce, thus generating a new emotional state (Rúas and García, 2018) that works to alter the current media ecosystem. Journalism must therefore face the new challenges and threats as various actors seek to impose their narrative regarding crises.

One tier below fake news, is so-called junk news. Without necessarily being false, it utilizes highly emotional and exaggerated language, with tabloid-like headlines that present biased and partisan information. They mimic the identity and style of legitimate sources in order to improve the credibility of their stories (García-Marín, Salvat-Martinrey, 2021).

This account of the story, sometimes alternative, also challenges the verification of facts as a democratic practice and the Western philosophical conceptions of absolute truth and “hostility towards plurality and contingency” (Arendt, 1993). Moreover, it poses a dilemma about the limits of information control and freedom of expression (Aguilera and Casero-Ripollés, 2018), thus reigniting a debate from the 1990s within UNESCO about the ethics of information or infoethics (Floridi, 2006) and current issues regarding data ethics.

Amongst the professional routines of journalism, information verification (Kovach and Rosenstiel, 2001) has historically been a fundamental one. Although this practice is far from new, the need for professional journalists to reinforce verification and rely on new strategies and new tools is even greater now, due to the emergence of social networks and the proliferation of fake news, which now spread at a faster pace.

For verification organizations, social networks, as tools in this task, are essential (Cinelli et al., 2020).; even more so considering that it is through these channels that a large amount of misinformation regarding the pandemic continues to flow (Salaverría et al., 2020; Sánchez-Duarte and Magallón-Rosa, 2020; Pérez-Dasilva et al., 2020) indicating that these are channels in which misleading information spreads more easily.

Although verification is essential in the information preparation process, as has been pointed out, journalistic quality cannot be understood at present without the concepts of transparency and innovation. The profession must continue to work on building a transparent model in which the newsrooms, together with the support of new tools and innovation, are capable of achieving this goal (López et al., 2018).

2. Automatic verification platforms and tools and AI

Artificial Intelligence (AI) can play a very important role in the effective fight against misinformation, as is the case of the so-called "good bots" (Flores-Vivar, 2019), which can help detect and fight against false news. Such is the case with "Fátima” (customized abbreviation for Fact Machine), a chatbot developed by the fact-checking site “Aos Fatos” in collaboration with Facebook, which helped combat fake news circulating through messenger during the Brazilian elections that led to the victory of Jair Bolsonaro. In addition to this tool, "Projeto Lupe", created by the Brazilian news agency Lupa, allowed users to request verified information mostly related to statements by electoral candidates.

Among these types of tools is also TruthBuzz, a program developed in collaboration with First Draft News, which helps reporters adopt compelling storytelling methods that improve the reach and impact of verification and help "inoculate" the public against false or misleading information. Facterbot, a Facebook Messenger “chatbot”, is a tool designed by a Spanish university student to stop far reaching false news ending up in users' inboxes. Fake News Detector also helps to automatically classify news based on its credibility, whilst Decoders, a section of the website of the newspaper Le Monde, evaluates the reliability of information sources through a specific search engine.

Credibility is determined, also automatically, based on the style of narration, through neural networks and stylometric characteristics that can pick up on the biased and effective vocabulary typical of fake news (Przybyla, 2020) as well as on the background of authors and associated sources (Sitaula et al., 2019).

In Spain there are some examples of the use of algorithms by political parties for the misuse of information extracted from personal data that show how robots are making progress as infomediators between politicians and citizens (Campos-Domínguez and García-Orosa, 2019), namely in the last Catalan (Salas, 2018) and Andalusian (Magallón et al., 2019) general and regional electoral campaigns. This has also been promoted by journalists and the media, as is the case with “Politibot”, the first news service on Telegram, in operation since the 2016 general elections. Similarly, within the academic field, there are examples such as the creation of a bot (unfaking_es) for monitoring information in the 2018 Andalusian elections through Twitter (Magallón et al., 2019), and the use of T-Hoarder, an open-source technology created primarily to automatically disseminate information from official sources, fact-checkers, journalists and the media.

In recent years, numerous automatic verification and fake news detection programs and tools (Atanasova et al., 2020) have been registered. Some of them are based on surface-level linguistic patterns, such as LIAR, which takes short statements tagged in various contexts of the Politifact.com platform, providing a detailed analysis report and links to documents (Yang Wang, 2017). Another example is UNTRUE NEWS (http://untrue.news) that uses ElasticSearch, an analytical search engine that draws from the Lucene library, providing results in real time (Woloszyn et al., 2020).

Research such as that of Vila-Márquez and Arce-García (2019) implements various techniques to keep track of an event or piece of news through the analysis of networks such as Twitter and the connections between users. This can be done in various ways. The first is by using the application of algorithms to determine the most followed authors or influencers (Gephi computer program), providing graphic representation through user nodes, their connections and links. Another way is through the use of text mining analysis, based on statistical framing techniques to determine which are the frames or main approaches (KH-Coder software). A stemming methodology, based on segmenting words, can also be implemented, eliminating biases and searching for morphological variants that can be grouped. Lastly, this monitoring can be achieved through the analysis of emotions to identify basic feelings (Fernández-Vallejo, 2018).

There are also conversation tracking programs on the Twitter API for the analysis of groups and social sentiments. Based on their location and time of publication, these include the tracking of emerging hashtags "#Socialdistancing" and "#workfromhome", as well as the online detection of cascading information with clickbait content (Sharma et al., 2020).

Some of these programs also perform information authenticity tests through initiatives based on Distributed Ledger Technologies (DKTs), which guarantee the origin, authenticity and traceability of the data to combat digital deception (Fraga-Lamas and Fernández-Caramés, 2020).

Likewise, as a result of the Coronavirus, a series of programs for the use of Artificial Intelligence, Machine Learning and mapping of resources have been developed to aid research efforts (Bullock et al., 2020). Equipped with high-precision machine learning classifiers, these tools can determine the validity of news based on word distributions and specific linguistic and stylistic differences in the first few sentences of an article (Qi Jia, 2019).

In recent times, new tools designed for the manipulation of images and video have emerged. "FaceForensis", is able to, among other things, attribute words to people that they did not say, something that is groundbreaking and unprecedented (Diakopoulos, 2018). Fortunately, photo verification instruments such as “TinEye”, Google's “Reverse Image Search”, and video tools, such as “Invid” or the “Reverso” site, have also been created to combat this type of manipulation.

Amazon has even incorporated into Alexa, the assistant of their smart speaker, an application that allows users to send questions to verifiers. Once again, this demonstrates the imbalance between the massive propagation of false information and the scarcity of resources, speed and efficiency to counteract its dissemination.

In fact, in recent years, digital technology and data science have been combined to facilitate the emergence of new forms of speech control and the establishment of powerful disinformation campaigns. This is possible through the manipulation of static or dynamic images that can be real, altered or created artificially (Gómez-de-Ágreda et al., 2021) using sophisticated techniques. The rise of a new generation of the so-called deep-fakes exemplifies the great complexity that image analysis entails in the post-truth era (Marzal-Felici, 2021).

In any case, AI can aid journalistic work in the production and selection process, in the fight against misinformation and the quest for objectivity. This technology can help avoid errors or ambiguities, identify information trends, aid in machine translation and also in the development of news recommendation systems (Manfredi Sánchez and Ufarte Ruiz, 2020; Túñez-López et al., 2019; Canavilhas, 2022).

The union of computational linguistics and journalism is evidenced in the combination of algorithms, data or knowledge from the social science field. This collaboration can bolster the public service responsibility attached to journalism, helping to classify and organize sources and overcoming the pressure for posting content. It also facilitates the automatization and speedy collection, analysis and verification of information. From automated to computational, algorithmic and robot journalism, this combination has resulted in different denominations that advance the search and structuring of relevant information.

3. Methodology

The objective of this study is to review the research on disinformation in Spain, by both examining the perspectives of the publications on the subject, and by creating of a map of groups and related research projects. To do this, a review of the modern literature on the subject was carried out. Then, a query was made in the database of the State Research Agency (AEI), in the section dedicated to grants, using the search keywords "disinformation”, “verification”, “fake news”, “factchecking”, “fact checking”.

Repetitions and projects not pertinent to the areas of Social Sciences, Law and Information and Communication Technologies were filtered out. 18 research projects, that were carried out and awarded between 2018 and 2021 and that included the aforementioned keywords, were analyzed. Also noted as study variables were the university that awarded the project, its budget, the area and subarea and the type of grant. Others such as the key and the object of study, the applied methodology and the context were investigated by the researchers of this study. The consultation was performed during the week of December 12 to 16, 2022, so the final approval that was communicated during those days, corresponding to 2021, is not included. However, the proof-of-concept projects for that year are included.

4. Results: Map of research projects on disinformation in Spain. Data from the State Investigation Agency

8 of the 18 projects extracted from the consultation on the AEI website come from the fields of Social Sciences -eight-, and from Information and Communication Technologies -five-. Projects in the fields Law and Education Sciences were also identified.

In the case of Social Sciences, five projects are identified in the Communication subarea, two in feminist, women's and gender studies and one in Sociology and Social Anthropology. Regarding the field of Information and Communication Technologies, the five projects belong to the sub-area of Computer Science and Information Technology.

In terms of keywords, the concept of "disinformation" is the one that appears most often across all eighteen areas and projects. Other important ideas also appear frequently, such as “social”, which is repeated in up to fourteen projects, “fakenews”, in ten cases, “factchecking” in eight, “media networks” in seven, “artificial intelligence” in six, "verification", "health", "imaginaries", "big data", "climate change" and "learning" in four and "digital literacy" in three.

By geographical area and university, the universities of the autonomous communities of Madrid, Catalonia and Galicia are the ones that produce the most projects on disinformation, followed by Andalusia and the Valencian Community.

Specifically, Madrid’s Carlos III University produced two projects, while The Autonomous University of Madrid, the National University of Distance Education and the Polytechnic University of Madrid all produced one. In Galicia, it is the University of Vigo that leads the way, with two projects within the field of of Social Sciences that relate to disinformation and verification. One focuses specifically on Communication, the other on Information and Communication Technologies.

By budget, the universities that lead the financing of projects related to disinformation are: Cádiz, with 139 thousand euros on a project, Vigo that totals 138.68 thousand euros across three projects, the UNED, at 94.86 thousand euros, the Polytechnic University of Madrid, at 92.44 thousand euros, the Carlos III University with 84.70 thousand euros and the Polytechnic Universities of Valencia and Barcelona with more than 70 thousand euros.

In summary, the total budget for projects awarded between 2018 and 2021 with a clear focus on disinformation was 1.06 million euros, representing an average of 106,446 euros per project.

Regarding the object of study, most of the projects here focused their research on the message, mainly on hate speech, political discourse, equality, health or immigration. A third of the studies focus on the audiences and the effects of disinformation, and only one investigation centers the study on the role of the sender.

Regarding the methodology proposed in the projects analyzed here, discourse analysis and content analysis predominate. In a third of the cases, the preferred methodologies are the analysis of networks and profiles, case studies, deep learning, surveys, big data, the study of professional routines and the analysis of the syllabus in the proposals framed within the area of Educational Sciences, which focus their commitment on media and digital literacy.

5. Discussion and Conclusions

According to the data consulted on the website of the State Investigation Agency, the map of national projects on disinformation and verification is concentrated in only six autonomous communities -Andalusia, the Valencian Community, Castilla-León, Catalonia, Galicia and Madrid-, and fourteen universities -UAM, UC3M, UCM, UB, UCA, UMA, USA, UV, UVIGO, UNED, UPM, UPV and UPF-.

By areas, Social Sciences, specifically Communication, Information and Communication Technologies predominate, although some projects were also found in Education Sciences and Law. This implies that, as indicated in the introduction, the study of disinformation is a growing research area, something that highlights the need to work between interdisciplinary teams in the medium and long term.

The keywords that predominate in the description of the projects are the following: social disinformation, fake news, fact-checking, artificial intelligence, verification, health, imaginaries, big data, climate change, 4C-based learning model and digital literacy. In this sense, it is striking that hardly any project includes ‘transparency’ among their keywords when, as indicated when citing López García et al. (2018), verification is directly linked to the construction of a transparent workflow for the newsrooms that prepare journalistic content.

When looking at the money granted to these projects, the universities that got the most funds are those of Cádiz (139 thousand euros), Vigo (138.68 thousand euros), UNED (94.86 thousand euros) and Carlos III (84.70 thousand euros). In total, the 18 projects resulting from the search for the keywords of the article add up to 1,064,460.00 euros, which means an average of 106,446.00 for each proposal.

By calls for papers, most of the projects fit into R+D+I (Generation of Knowledge and Research Challenges), except for three that are from R+D -Excellence / Generation of Knowledge-, and one, from the University of Vigo (The fight against misinformation and value criteria in electoral debates on television and digital models: verification platform and blockchain) that first into proof of concept.

In conclusion, the study of the concept of disinformation and false news has been attracting academia for some time. The number of investigations and projects focusing on this topic are increasing. That disinformation’s ability to deceive and misinform has become so evident has fostered greater research and investigation into the phenomenon. To gauge this growth, it would be interesting to complete this study with the number of projects that were awarded in international and regional calls.

In addition, it would be interesting to analyze the data available on the State Investigation Agency’s page regarding the total number of projects that were presented versus those that were finally granted.

Author’s contribution

José Rúas-Araújo: Conceptualization, Investigation, Methodology, Validation, Verification, Visualization, Writing- original draft and Writing-revision and editing. Francisco-Javier Paniagua-Rojano: Conceptualization, Investigation, Methodology, Validation, Verification, Visualization, Writing-original draft and Writing-revision and editing. All authors have read and agree to the published version of the manuscript. Conflicts of interest: The authors declare that they have no conflict of interest.

Funding

This study is part of the research activities promoted through the research project PID2021-124293OB-I00, of the State R+D+I Program for the realization of Knowledge Generation projects 2021, "Map of Disinformation in Regional and Local Entities of Spain and its Digital Ecosystem (FAKELOCAL)".

Furthermore, it is part of the activities promoted through the transfer project PDC2021-121720-100, the State R+D+I Program to perform "Proof of Concept", the State Challenges of Society Program, the State Program for Scientific, Technical and Innovation Research 2017-2020 "Fight against disinformation and value standards in electoral debates on TV and digital media: verification platform and blockchain (DEBATrue)".

References

Author notes

* Professor of Political Communication at the School of Social Sciences and Communication

** Professor of Corporate and Institutional Communications

Additional information

Translation to English : Karina Rivero Conde (Xurxo Vaca).

To cite this article : Rúas-Araújo, José; & Paniagua-Rojano, Francisco-Javier. (2023). Approximation to the map on disinformation and verification research in Spain: state of affairs. ICONO 14. Scientific Journal of Communication and Emerging Technologies, 21(1). https://doi.org/10.7195/ri14.v21i1.1987

ICONO 14, Revista de comunicación y tecnologías emergentes

ISSN: 1697-8293

Vol. 21

Num. 1

Año. 2023

Approximation to the map on disinformation and verification research in Spain: state of affairs

José Rúas-Araújo ¹, Francisco-Javier Paniagua-Rojano ²