SODAS Data Discussion II (Fall 2026)
SODAS Data Discussion with Sergey Belmestnykh and Benjamin Kohler.
Copenhagen Center for Social Data Science (SODAS) aspirers to be a resource for all students and researchers at the Faculty of Social Sciences. We therefore invite researchers across the faculty to present ongoing research projects, project applications or just a loose idea that relates to the subject of social data science.
Two researchers will present their work. The rules are simple: Short research presentations of ten minutes are followed by twenty minutes of debate. No papers will be circulated beforehand, and the presentations cannot be longer than five slides.
Discussion 1
Presenter: Sergey Belomestnykh
Title: Words as Species: Mapping Research Communities on AI and Sustainability
Abstract:
Rather than reaching for the usual toolkit, LDA or BERTopic for topic modelling, or a full LLM-based pipeline, we took a detour through ecology to investigate the role of AI in SDG-related research. We systematically reviewed the most-cited literature at that intersection and ran hierarchical clustering, detrended correspondence analysis, and indicator species analysis. This talk also intends to showcase the versatility of data science and underscores the importance of interdisciplinary that is, how knowledge from one discipline can be applied in the context of another. The study is conducted by Sergey Belomestnykh and Jorge Gustavo Rodríguez Aboytes.
Discussion 2
Presenter: Benjamin Kohler
Title: AI Guidance and Admission Choices: Experimental Evidence
Abstract:
In high-stakes decisions like college choice, information frictions can impede effective decision-making even when reliable information is publicly available. We study whether conversational AI can reduce these frictions in a preregistered field experiment with young adults considering higher-education applications in Denmark (N=2,187). Participants were randomly assigned to a domain-specific AI chatbot -- designed for information retrieval and program comparison based on individual student profiles -- or to the official Danish education-guidance website. Participants rate the chatbot as more useful and spend more time with it than with the website control. The chatbot produces smaller gains in perceived informedness than the website, yet greater improvements in objective knowledge about admissions cutoffs. Access to the chatbot increases the rate in which students revised their planned programme ranking, in particular by introducing first-choice programmes that were not in their initial list. These results suggest that conversational AI can affect consequential choices by improving personalized information access, in particular by expanding users' consideration sets.
