In recent years, there has been a strong public interest in scientific knowledge. Ordinary citizens as well as politicians, celebrities, journalists and other stakeholders have actively participated in discussions about scientific knowledge and science itself through online platforms. Wider access to and interest in scientific information as such can be seen as a positive development, especially in the era of generative AI where reliable and verifiable information is critical in order to discern misinformation and facts. However, the typically rather informal and sometimes decontextualized science-related discourse on the Web may result in (deliberate or accidental) oversimplification, misrepresentation and instrumentalization of scientific knowledge. This may facilitate misinformation, hamper decision-making by politicians, have a negative impact on citizens’ trust in science and ultimately increase the polarization of society. Investigating science-related online discourse is therefore a major challenge for the scientific community today. However, science-related Web discourse has different characteristics than general Web discourse which reduces the effectiveness of general-purpose computational methods. Also, online discourse tends to cross various platforms, e.g. people may refer to scientific claims from publications, news portals or YouTube videos in posts on the platform X.
Our workshop shall serve to advance interdisciplinary research concerning science-related discourse on the Web by providing a platform for the exchange of research and resources including computational methods for platform-specific and cross-platform data extraction and analysis as well as social scientific insights into science-related communication on the Web and possible societal effects.
We invite contributions from Computer Science, Computational Social Science, Communication Science, Science Communication, Media and Communication Studies, Information Science, Computational Linguistics, and related fields.
Workshop Themes
The workshop focuses on two main themes:
-
- Methodological challenges related to the analysis of science-related discourse on the Web, including data acquisition and processing
- Insights concerning science-related discourse on the Web, e.g., its characteristics, evolution, and impact
Topics of interest within these themes include the following:
-
- (Cross-platform-) crawling approaches for science-related discourse
- Methods for the detection and filtering of science-related online discourse data
- Issues and methods related to tracking/linking of users and messages within and across different platforms (e.g., X, Bluesky, Mastodon, Threads)
- Practical/legal/ethical issues concerning data access
- Detection of arguments, claims, evidence, sources, or stances in science-related online discourse data
- Classification of scientific claims w.r.t. verifiability, credibility, or veracity (including distinguishing different types of misinformation such as oversimplification)
- Analysis of science-related discourse on social media platforms and in online news
- Assessment of the expertness of social media users (e.g., scientist, expert, lay person)
- Classification of sources w.r.t. credibility, political leaning, and other biases
- Usage of social media platforms (e.g., X, Bluesky, Mastodon, Threads) by different user groups
- Usage of memes in science-related discourse on social media
- Analysis of LLM-generated text in science-related discourse
- Usage of preprints and open access publications in science-related Web discourse
SDW2026 