Articles
Permanent URI for this collectionhttps://www.weizenbaum-library.de/handle/id/1111
Collection of articles, reviews, preprints
Browse
6 results
Search Results
Item LGDE: Local Graph-based Dictionary Expansion(2025) Schindler, Juni; Jha, Sneha; Zhang, Xixuan; Bühling, Kilian; Heft, Annett; Barahona, MauricioWe present Local Graph-based Dictionary Expansion (LGDE), a method for data-driven discovery of the semantic neighbourhood of words using tools from manifold learning and network science. At the heart of LGDE lies the creation of a word similarity graph from the geometry of word embeddings followed by local community detection based on graph diffusion. The diffusion in the local graph manifold allows the exploration of the complex nonlinear geometry of word embeddings to capture word similarities based on paths of semantic association, over and above direct pairwise similarities. Exploiting such semantic neighbourhoods enables the expansion of dictionaries of pre-selected keywords, an important step for tasks in information retrieval, such as database queries and online data collection. We validate LGDE on two user-generated English-language corpora and show that LGDE enriches the list of keywords with improved performance relative to methods based on direct word similarities or co-occurrences. We further demonstrate our method through a real-world use case from communication science, where LGDE is evaluated quantitatively on the expansion of a conspiracy-related dictionary from online data collected and analysed by domain experts. Our empirical results and expert user assessment indicate that LGDE expands the seed dictionary with more useful keywords due to the manifold-learning-based similarity network.Item Veiled conspiracism: Particularities and convergence in the styles and functions of conspiracy-related communication across digital platforms(2025) Bühling, Kilian; Zhang, Xixuan; Heft, AnnettDigital communication venues are essential infrastructures for anti-democratic actors to spread harmful content such as conspiracy theories. Capitalizing on platform affordances, they leverage conspiracy theories to mainstream their political views in broader public discourse. We compared the word choice, language style, and communicative function of conspiracy-related content to understand its platform-dependent differences and convergence. Our cases are the conspiracy theories of the New World Order and Great Replacement, which we analyzed on 4chan/pol/, Twitter, and seven alternative US news media longitudinally from 2011 to 2021. The conspiracy-related texts were comparatively analyzed using a multi-method approach of computational and quantitative text analyses. Our results show that conspiracy narrations are increasingly present in all venues. While language differs vastly between platforms, we observed a style convergence between Twitter and 4chan. The results show how more coded language veils the spread of racist and antisemitic content beyond the so-called dark platforms.Item Challenges of and approaches to data collection across platforms and time: Conspiracy-related digital traces as examples of political contention(2024) Heft, Annett; Bühling, Kilian; Zhang, Xixuan; Schindler, Dominik; Milzner, MiriamTaking the example of conspiracy-related communication online as one form of contentious politics, this study examines the data collection challenges for multidimensional comparative research across platforms, time, and cultural embeddings. It compares the architectures and features relevant to data collection, access regimes, and use cultures for a set of digital platforms and communication venues. Differentiating between actor- and content-based strategies, this study discusses the potentials and limitations of these approaches, considering differences in platforms, temporal dynamics, and cultural embeddings as well as several layers of equivalence. The discussion highlights crucial insights into designing data collection strategies in multidimensional comparative studies.Item Mapping a Dark Space: Challenges in Sampling and Classifying Non-Institutionalized Actors on Telegram(2023) Jost, Pablo; Heft, Annett; Bühling, Kilian; Zehring, Maximilian; Schulze, Heidi; Bitzmann, Hendrik; Domahidi, EmeseCrafted as an open communication platform characterized by high anonymity and minimal moderation, Telegram has garnered increasing popularity among activists operating within repressive political contexts, as well as among political extremists and conspiracy theorists. While Telegram offers valuable data access to research non-institutionalized activism, scholars studying the latter on Telegram face unique theoretical and methodological challenges in systematically defining, selecting, sampling, and classifying relevant actors and content. This literature review addresses these issues by considering a wide range of recent research. In particular, it discusses the methodological challenges of sampling and classifying heterogeneous groups of (often non-institutionalized) actors. Drawing on social movement research, we first identify challenges specific to the characteristics of non-institutionalized actors and how they become interlaced with Telegram’s platform infrastructure and requirements. We then discuss strategies from previous Telegram research for the identification and sampling of a study population through multistage sampling procedures and the classification of actors. Finally, we derive challenges and potential strategies for future research and discuss ethical challenges.Item Pandemic protesters on Telegram: How platform affordances and information ecosystems shape digital counterpublics(2023) Bühling, Kilian; Heft, AnnettThis study analyzes how platform affordances, their appropriation by movement actors, and these actors’ leveraging of information ecosystems—in combination—helped form a digital counterpublic during the COVID-19 pandemic. It draws on public communication data sent by more than 300 Telegram channels and group chats affiliated with the Querdenken movement over a 2-year period, and combines automated and manual text classification with network analysis. The study demonstrates how Telegram afforded connective and collective action in distinct ways that reflected the movement’s organizational structure and aims, as well as the impact of individual information-sharing on the process of movement-building itself. Accounting for time-dependent dynamics, the study also found that different parts of the counterpublic latched onto and sustained distinct information ecosystems to articulate their claims and mobilize contentious action.Item Measuring the diffusion of conspiracy theories in digital information ecologies(2022) Heft, Annett; Bühling, KilianDigital platforms and media are fertile breeding grounds for disinformation and conspirational views. They provide a variety of communication venues for a mixed set of actors and foster the diffusion of content between actor groups, across platforms and media, and across languages and geographical spaces. Understanding those diffusion processes requires approaches to measure the prevalence and spread of communicative acts within and across digital platforms. Given the increasing access to digital data, computational methods provide new possibilities to capture this spread and do justice to the interrelated nature and hybridity of online communication. Against this background, the paper focuses on the spread of conspiracy theories in digital information ecologies. It provides a review of recent methodological approaches to measuring conspiracy-related content online regarding the (a) prevalence and (b) diffusion of conspiracy theories. To that end, the paper differentiates between social network analysis approaches and computational techniques of automated text classification. It further discusses how far these and related computational approaches could facilitate studying the diffusion of conspiracy theories across different actor types, languages, topics and platforms. In doing so, it takes the specific nature of online communication and challenges in the field of conspiracy-related content into account.