Search

This conceptual paper explores the role of communication around data practices of Big Tech companies. By critiquing communication practices, we argue that Big Tech platforms shape users into data subjects through framing, influencing behaviour, and the black-boxing of algorithms. We approach communication about data from three perspectives: (1) current data communication constructs reductive data identities for users and contributes to the colonization of daily routines; (2) by strategically deploying the black box metaphor, tech companies try to legitimize abuses of power in datafication processes; (3) the logic in which communication is mediated through the interfaces of Big Tech platforms is normalizing this subjectification. We argue that critical data literacy can foster individual resilience and allows users to resist exploitative practices, but this depends on transparent communication. The opposite seems standard among tech companies that obfuscate their data practices. Current commercial appropriations of data ethics need to be critically assessed against the background of increasing competition in the digital economy

Artificial Intelligence’s usage and regulation has become a highly debated topic that unveils the political, economic and technoscientific nuances between and within cultures. AI has garnered considerable attention in the public debate, being thought of and envisioned differently by national governments, technology corporations, media outlets, academia and citizenship. While previous research has focused on finding these imaginaries in support of technological advances in national policy-making, media, and the social media landscape, online communities remain largely unexplored. The study of online communities has become an increasingly challenging research subject that comprehends a debate atmosphere that encourages the exchange of ideas, where stakeholders engage, challenge, and negotiate dominant AI discourses and imaginaries. Through a mixed-methods frame analysis, the present thesis explores and critically assesses how online communities surrounding AI-generated livestreams debate and frame AI sociotechnical imaginaries in comparison to media outlets and national policies. Unpacking these sociotechnical imaginaries has revealed the dominant presence of technologically deterministic discourses, but also the blind spots in the discussion, that portray AI as an inevitable and transformative agent in society. These discourses are mainly characterized by hiper-binary perspectives over the usage and envisioning of AI regarding its transformative potential, often overshadowing ethical considerations and critical perspectives, and augmenting AI’s capabilities even beyond its actual limits.

The launch of ChatGPT, an AI chatbot, has ignited a profound discussion on the future of artificial intelligence. With its impressive capabilities and wide-ranging applications, ChatGPT has garnered both praise and criticism, attracting 1 million users within only 5 days. While scholars and experts delve into the long-term implications of these technologies, people have shared their experience with the newest technologies on social media, including Twitter, a platform known for its microblogging and widespread discussions between users. Recognizing the significance of this recent development in the field of large language models, this study sets out to explore the perceptions of ChatGPT by analysing a vast dataset of 2.6 million Twitter posts using natural language processing techniques. By examining emphasis frames, sentiments, and the key actors involved, this research aims to shed light on the nature of the debates surrounding ChatGPT.

The COVID-19 pandemic raised questions about trust in journalism and the quality of news reporting during societal crises. While journalists and media professionals frequently offered critical reflections based on personal experiences and observations, computational methods are not widely used to support these evaluative processes. We aim to extend the conversation on metajournalistic discourse by considering the inclusion of empirical methods for monitoring journalistic practices. By disclosing our findings about Dutch news media’s corona reporting between 2019 and 2022, we demonstrate how computational methods for content analyses of news texts can yield empirically informed insights into different facets of journalistic performance. The corpus includes 106,616 corona-related articles from national and regional newspapers in the Netherlands. We deployed text analytical methods such as topic modelling and named entity recognition to explore Dutch corona reporting in respect to different normative criteria (informing, monitoring, offering platforms for discussion and opinion, interpretation, analysis, and setting public agendas). The study was requested by a large Dutch newspaper to receive a systematic-empirical analysis of journalistic practice for self-evaluation. We argue that computational methods combined with qualitative analyses can stimulate dialogue and critical reflection among news media professionals.

Media framing of organizational crises is an important factor to consider in crisis communication since it can shape stakeholders' perceptions of organizations and discussions in the public sphere. This takes place in complex media ecologies where public communication happens at a large scale, both in the news and on social media. Here, computational methods offer new venues for analysing media framing in flux throughout the crisis life cycle. Especially methods for automated content analysis can quickly and efficiently reveal what media frames emerge in a crisis context and how they change over time across different channels and platforms. The present study showcases the benefits of such methodological approaches by critically exploring the example of the data breach at the national municipal health service in the Netherlands. Using computational methods for media frame analysis on news texts (N1 = 519) and social media postings (N2 = 2986), this article reconstructs how the incident was perceived throughout four crisis stages (build-up, outbreak, chronic stage, termination). The article critically discusses the relevance of researching media framing empirically with emphasis on the benefits but also limitations of computational approaches. It concludes with some general pointers for crisis researchers interested in such methods as well as their implications for practitioners in the field.

The news media play an important role to assist citizens in becoming informed about crises, which allows them to evaluate the causes and responsibilities. When publishing news, news media can utilise frames to influence the perceptions and impressions of their audiences. Previous research has manually analysed the used frames regarding the coverage of the Dutch nitrogen crisis, but did so for a short period of time or did not cover the differences between the analysed outlets. Therefore, journalistic practices, regarding this crisis, have not yet been thoroughly analysed, which shows a gap in the research. The current research tries to fill this gap by analysing longitudinal shifts regarding the coverage of the Dutch nitrogen crisis. The aim of this study is to determine if computational methods can be used to determine imbalances and biases in framing practices, in order to monitor journalistic practices. To research this, 9,374 articles from June 2019 until March 2024 were extracted and analysed using Natural Language Processing techniques. Topic modelling and Named Entity Recognition were applied to perform a framing analysis on the dataset, which were visualised using knowledge graphs. The results show shifts regarding the topics and entities, with the topics being quite diverse and the entities being mostly framed politically. Lastly, the knowledge graphs showed to be successful in capturing the context of the topics and entities, while also indicating a difference between the frames that outlets applied. The results show that these methods can be used to develop a tool for analysing journalistic practices. Development of this tool should start after addressing the limitations, while the current study did not implement sentiment regarding the topics and only allowed for one topic per article. Additionally, it would be good practice to extend the case beyond the nitrogen crisis, to validate the findings of this study. The developed tool can support news organisa

Methodologies and data are not currently instantly extractable from digital humanities papers. Biology and medicine have a tradition of using named entity recognition to tackle this lacuna. In this thesis, a named entity recognition model, trained on a sample of 197 relevant annotated five-sentence windows extracted from a broad corpus of 692 English-language Digital Humanities Quarterly articles to label entities as either ‘methodology’, ‘method’, ‘tool’, ‘task’ or ‘data’, is employed to multifocally read digital humanities methodologies. Multifocal reading practices are found to be promising, but performance and external validity of the named entity recognition model are lacking.

Public discourses have become sensitive to the ethical challenges of big data and artificial intelligence, as scandals about privacy invasion, algorithmic discrimination, and manipulation in digital platforms repeatedly make news headlines. However, it remains largely unexplored how exactly these complex issues are presented to lay audiences and to what extent news reporting—as a window to tech debates—can instil critical data literacy. The present study addresses this research gap and introduces the concept of “data risks”. The main goal is to critically investigate how societal and individual harms of data-driven technology find their way into the public sphere and are discussed there. The empirical part applies a mixed methods design that combines qualitative and automated content analyses for charting data risks in news reporting sampled from prominent English-speaking media outlets of global reach. The resulting inventory of data risks includes privacy invasion/surveillance, data bias/algorithmic discrimination, cybersecurity, and information disorder. The study posits data risks as communication challenges, highlights shortcomings in public discussions about the issue, and provides stimuli for (practical) interventions that aim at elucidating how datafication and automation can have harmful effects on citizens.