Scholars present at ICA 2024

Authors: Alfred Hermida, Felix M. Simon

Institutions: A. Hermida, Journalism, U of British Columbia, Vancouver, British Columbia, Canada; F.M. Simon, Oxford Internet Institute, U of Oxford, Oxford, UK

Abstract: This presentation considers the deployment of AI in newsrooms based on a case study of Sophi, an algorithmic recommendation engine developed by Canada’s leading quality newspaper, The Globe and Mail. Sophi automates and optimizes most publishing decisions, and can be used to encourage subscriptions through a dynamic paywall. Recognized with international industry awards, the AI engine has been largely adopted by media groups to manage local news websites at scale. This presentation suggests that the experience of Sophi provides a model to analyze the adoption of AI technologies at an organizational level in journalism. In this model, news publishers would be keen to adopt AI in cases that combine low reputational risk with high efficiency and cost-savings. There would be a greater degree of resistance and less adoption in cases that combine high reputational risk with low efficiencies and cost-savings. The implications for journalism and audiences will be explored.

Authors: Sibo Chen, Nicole Blanchett, Charles Davis

Institutions: S. Chen, School of Professional Communication, Toronto Metropolitan University, Toronto, Ontario, Canada; N. Blanchett, The Creative School, Ryerson U, Toronto, Ontario, Canada; C. Davis, Faculty of Communication & Design, Ryerson U, Toronto, Ontario, Canada;

Abstract: This presentation considers the deployment of AI in newsrooms based on a case study of Sophi, an algorithmic recommendation engine developed by Canada’s leading quality newspaper, The Globe and Mail. Sophi automates and optimizes most publishing decisions, and can be used to encourage subscriptions through a dynamic paywall. Recognized with international industry awards, the AI engine has been largely adopted by media groups to manage local news websites at scale. This presentation suggests that the experience of Sophi provides a model to analyze the adoption of AI technologies at an organizational level in journalism. In this model, news publishers would be keen to adopt AI in cases that combine low reputational risk with high efficiency and cost-savings. There would be a greater degree of resistance and less adoption in cases that combine high reputational risk with low efficiencies and cost-savings. The implications for journalism and audiences will be explored.

Presenter: Michelle Riedlinger

Authors: Michelle Bartleman, Elizabeth Dubois

Institutions: M. Bartleman, Communication, University of Ottawa, Ottawa, Ontario, Canada; E. Dubois, Communication, U of Ottawa, Ottawa, Ontario, Canada

Abstract: This study builds on Danzon-Chambaud’s 2021 systematic review to bring up to date and expand on the state of automated journalism scholarship. Thanks to increasingly abundant and openly available data, combined with artificial intelligence (AI)-enabled tools, over the past decade there has been a steady increase of news outlets publishing machine-generated textual news content, known most commonly as automated journalism. Danzon-Chambaud (2021) identified 33 empirical studies published in English between 2005 and 2020. However, the amount of research has since increased dramatically, with the number of articles doubling in the past two years. This study expands the previous review to account for a growing number of related terms, includes theoretical and conceptual studies, and aggregates critical issues.. Furthermore, thanks to its Canadian origin, this review also considers French language scholarship, further expanding our global knowledge of this quickly developing research area.

Authors: Ned Watt, Michelle Riedlinger

Institutions: N. Watt, QUT, Brisbane, Queensland, Australia; M. Riedlinger, Creative Industries, Queensland U of Technology, Brisbane, Queensland, Australia

Abstract: Fact checkers recognise that Generative AI (GenAI) technologies can create large volumes of false, synthetic visual and textual content, and generate inauthentic social media user engagement. However, GenAI technologies can also positively benefit fact checkers. This presentation draws on interviews with fact checkers in sub-Saharan Africa, Latin America and Australia in the months after popular and competitive LLM products, including the ChatGPT API for developers and Google’s Bard were released. We found that fact checkers imagine GenAI technologies in mainly “helper” roles, supporting them with problematic narrative tracking, streamlining large-scale verification (e.g. checking parliamentary reports and online video content), identifying expert sources, and responding to previously debunked “zombie” claims. While fact checkers imagined using GenAI technologies to improve audience engagement, mainly through repurposing existing content, they were yet to see roles for themselves in shaping or co-creating GenAI technologies with the communities they serve or transforming their practices.

Authors: Elizabeth Dubois

Institutions: E. Dubois, Communication, U of Ottawa, Ottawa, Ontario, Canada

Abstract: Social media influencer marketing is being integrated into political campaigning around the world wherein influencers are recruited to send party messages to their followers. Political communication studies have long relied on the concept of an opinion leader, who uses social pressure and support to personally influence their associates, to explain this type of bridging role. But influencers may have parasocial, rather than close personal, relationships with their followers; and what makes an influencer influential can vary by context, platform, size of audience, and other factors. We collect descriptions of political influencer programs in North America from 26 marketing and advertising firms and 5 political and advocacy organization. Using definitions on their websites and other public communications, our content analysis will highlight common themes in how influencers and their political roles are conceptualized. We discuss how this empirical evidence aligns and diverges from understandings of social media influencers more broadly as well as political opinion leaders more specifically.

Presentation type: Extended Abstract: This is not a student led paper.

Authors & Institutions: K. Esau, School of Communication, Digital Media Research Centre (DMRC), Queensland U of Technology, Brisbane, Queensland, Australia; H. Meyer, Journalism and Communication Studies, U of Hamburg, Bremen, Bremen, GERMANY; M. Farjam, Lund U, Lund, SWEDEN; A. Bruns, Digital Media Research Centre, Queensland U of Technology, Kelvin Grove, Queensland, Australia

Secondary Category: Computational Methods

Abstract: This empirical study investigates the media framing of climate protests in Australia and Germany. Media frames serve as powerful tools for shaping public perceptions of complex social issues. While previous research has focused on the framing strategies employed by climate activists, we are examining how news media outlets themselves frame the climate protests within different political contexts. This is particularly relevant as climate protests have become a focal point in contentious public debates in recent years. Employing a mixed-methods approach, we integrate qualitative and computational methods to identify frames used in news stories about climate protests. This study pioneers a dual-language approach, encompassing both English and German, thereby enriching frame analysis in political communication research. Employing a comparative approach, we consider the political leanings of media outlets, different types of protests, and the influence of political and media systems in two countries, assessing the impact of these factors on the framing of climate protests. The results reveal that right-leaning media outlets frame the protests and activists more negatively, defining them as a problem to societal cohesion rather than a solution to climate change. Regarding the relationship between media, political context, and climate protest framing, the results show that the Australian media landscape exhibits less diverse framing compared to Germany. We discuss the results in the context of news media polarisation and public opinion formation.

Presenter: Katharina Esau

Author & Institution: A. Azmi, Digital Media Research Centre, Queensland U of Technology, Brisbane, Queensland, Australia

Secondary Category: Feminist Scholarship

Abstract: This study problematises the concept of the global movement when it comes to online movement like #MeToo, which is often said to “spread worldwide” after the viral use of #MeToo hashtag in the US. The word “global” poses ambiguous meaning when trying to understand the process of the growth of a movement, while the extent to which global and local movement influence each other remains complex. Different sociopolitical and cultural specificities mean that #MeToo movement has different characteristics in many different places.

Indonesia, as the world’s most populous Muslim nation, had not seen the significance of the #MeToo movement in 2017, despite being dubbed the world’s social media capital—due to its high number of users and the young population. Online movement against sexual violence progressed in later years, particularly during the controversies about the now legislated Law on Sexual Violence Crimes (UU TPKS), which took six years of deliberation in the legislative People’s Representative Council (DPR).

Focusing on two topics about sexual violence, I present how conversations about sexual violence in Indonesia emerged despite the challenges and different characteristics from the #MeToo movement in the Global North. The first topic focus on the social media conversation about a well-known singer who posted her experience of sexual harassment on Instagram in 2018, around the time the #MeToo hashtag was used widely in the Global North. The second topic focus on the controversies about the Elimination of Sexual Violence bill (P-KS bill), from 2019 when the bill gained traction on social and mainstream media, to 2022 when it was passed as the TPKS Law. I apply social media analysis of post activity data collected from Facebook, Instagram, and platform formerly known as Twitter. I present data quantitatively to reveal peak activities, related themes, and prominent actors in the conversations, and analyse the data qualitatively to explain how the movement unfolded and expanded discussions about sexual violence. This study demonstrates Indonesia’s movement against sexual violence and reflects on the dialogical or resistive relationship with the global #MeToo movement and improve understanding of the diverse struggles of discussing sexual violence despite the claim of the global #MeToo movement’s success.

Presenter: Alia Azmi

File Upload: [ICA 24 – Global and local movement relationship 21.pdf]

Authors: Daniel Angus, Stephen Harrington, Axel Bruns, Phoebe Matich, Nadia Jude, Edward Hurcombe, Ashwin Nagappa

Institutions: D. Angus, School of Communication, Queensland U of Technology, Kelvin Grove, Queensland, Australia; S. Harrington, Queensland U of Technology, Brisbane, Queensland, Australia; A. Bruns, Digital Media Research Centre, Queensland U of Technology, Kelvin Grove, Queensland, Australia; P. Matich, N. Jude, A. Nagappa, Digital Media Research Centre, Queensland University of Technology, Brisbane, Queensland, Australia; E. Hurcombe, School of Media and Communication, RMIT U, Melbourne, Victoria, Australia;

Abstract: This paper focusses on a computationally-assisted analysis of the broad topics of discussion in 954 Facebook pages and groups that are most strongly implicated, and highly visible, in the sharing of links to 2314 high-profile ‘fake news’ websites, between 2016 and 2022. Our approach allows us to, first, understand what else (beyond problematic content) users share in these spaces and, second, how a specific topic of interest may play a role in driving users to share mis- and disinformation.

Presentation type: Extended Abstract: This is not a student led paper.

Authors & Institutions: M. Riedlinger, Creative Industries, Queensland U of Technology, Brisbane, Queensland, Australia; V. García-Perdomo, Journalism, Universidad de La Sabana, Bogota, Cundinamarca, COLOMBIA; N. Watt, QUT, Brisbane, Queensland, AUSTRALIA; M.F. Orjuela Alabarracin, Faculty of Creative Industries, Education and Social Justice, Queensland U of Technology, Dakabin, Queensland, AUSTRALIA; M. Joubert, Centre for Research on Evaluation, Science and Technology (CREST), University of Stellenbosch, Stellenbosch, Stellenbosch, SOUTH AFRICA; S.X. Montaña-Niño, Journalism program, University of Melbourne, Melbourne, Victoria, AUSTRALIA; E. van Zuydam, CREST, Stellenbosch U, Pretoria, Gauteng, SOUTH AFRICA;

Secondary Category: Political Communication

Abstract: This study advances the debate about the institutionalization of fact checking by outlining the emergence of alternative normative values associated with the current verification practices undertaken by global fact checking organizations. Drawing on the boundaries framework and jurisdictional reflections elicited by Graves (2018), and inspired by news values theory (Galtung & Ruge, 1965; Harcup & O’Neill, 2017; Shoemaker & Reese, 2014), and subsequent social media news works discussed by Hermida (2012; 2019) and García-Perdomo and colleagues (2017), this article lays the foundation for developing a similar model of values for fact-checking. We foreground a set of 6 fact-checking values, including truthfulness based on the documents produced by these organisations, online content produced in 2021 and a body of 36 semi-structured interviews conducted across Meta-affiliated organizations.

Presenter: Silvia Ximena Montaña-Niño

Presentation type: Extended Abstract : This is a student-led paper.

Authors & Institutions: N. Watt, QUT, Brisbane, Queensland, Australia; M. Riedlinger, Creative Industries, Queensland U of Technology, Brisbane, Queensland, Australia; M.F. Orjuela Alabarracin, Faculty of Creative Industries, Education and Social Justice, Queensland U of Technology, Dakabin, Queensland, Australia; M. Joubert, Stellenbosch U, Stellenbosch, South Africa; V. García-Perdomo, Journalism, Universidad de La Sabana, Bogota, Cundinamarca, Colombia; E. van Zuydam, CREST, Stellenbosch U, Pretoria, Gauteng, South Africa; S.X. Montaña-Niño, Journalism program, University of Melbourne, Melbourne, Victoria, Australia

Abstract: This study contributes to the emerging body of work on AI-assisted fact checking by specifically investigating how fact checkers imagine and make use of emerging GenAI technologies in their day-to-day work, particularly Large Language Models (LLMs). We particularly focus on platform-supported fact checkers in Southern hemisphere regions and the language, socio-political contexts and equity differences that exist when compared with platform-supported fact checkers in North America and Europe.

Presenter: Ned Watt

Authors & Institutions: A. Fleerackers, Interdisciplinary Studies, Simon Fraser U, Vancouver, British Columbia, Canada; L.L. Moorhead, Journalism, San Francisco State U, San Francisco, California, United States; L. Maggio, Medicine, Uniformed Services University, Bethesda, Maryland, United States; M. Riedlinger, Creative Industries, Queensland U of Technology, Brisbane, Queensland, Australia; J. Alperin, School of Publishing, Simon Fraser U, Vancouver, British Columbia, Canada

Secondary Category: Environmental Communication

Abstract: The medialization of science is often used to describe the ways in which scientists bend towards the norms, values, practices, or “logics” of journalism. Yet, the concept of medialization is theoretically bidirectional and could also be used to understand how journalists adopt the “logics” of scientists. This study sheds light on this overlooked aspect of medialization through a qualitative analysis of 19 interviews with health and science journalists about the norms, practices, and criteria they use to select, verify, and communicate research. It presents a structured framework for analyzing the impact of medialization on journalists (rather than scientists) and demonstrates that such analyses can provide important insights into the nature and implications of journalists’ adoption of science logic—what we might call the “scientization” of journalism.

Presenter: Alice Fleerackers