Options
Viktor Suter
Last Name
Suter
First name
Viktor
Email
viktor.suter@unisg.ch
ORCID
Phone
+41 71 224 3474
Now showing
1 - 5 of 5
-
PublicationCOVID-19's (mis)information ecosystem on Twitter: How partisanship boosts the spread of conspiracy narratives on German speaking Twitter( 2021)
;Shahrezaye, Morteza ;Steinacker, LéaType: conference paper -
PublicationFacing the public: A cross-national analysis of social norms and communication about facial recognition technologies( 2020-05)
;Steinacker, Léa ;Kostka, Genia ;Guo, DanqiType: conference paper -
PublicationType: conference paper
-
PublicationUsing GPT-4 for Text Analysis: Insights from English and German Language News Classification Tasks( 2024-06-01)Large language models are rapidly becoming an essential tool for social scientists. In particular, they have the potential to completely change the way researchers approach text analysis. In this study, we use the GPT-4 model to classify the content of newspaper articles and assess their sentiment. To do this, we collect headlines and leads from U.S. and German newspapers (n = 1,629) on how generative AI is represented in major news media outlets in both countries and inductively develop coding instructions based on this data. We then feed the data and instructions to GPT-4 and a human coder to compare their outputs and assess validity. We find that the coding procedure is highly reliable, with substantial to near perfect agreement between the human coder and GPT-4. We also find that reliability decreases for more complex constructs and is modestly lower for classification tasks performed in German than in English. Based on this analysis, we argue that LLMs offer powerful new approaches to text analysis that cross methodological divides between qualitative and quantitative approaches to empirical text analysis.Type: conference contributionJournal: REAL-Info 2024: First Workshop on Reliable Evaluation of LLMs for Factual InformationDOI: 10.36190/2024.31
-
PublicationDigital Contact Tracing in Switzerland: A Computer-Assisted Qualitative Analysis( 2023)Public administrations often face significant challenges when adopting new digital technologies. The introduction of contact tracing during the pandemic provided an opportunity to assess the challenges encountered by public authorities when implementing digital technologies. We conducted semi-structured qualitative expert interviews with representatives of federal and cantonal agencies, as well as the Swiss National COVID-19 Science Task Force, to explore their opinions and viewpoints on the implementation of SwissCovid, a contact tracing application deployed by the Swiss government. The interview data were analyzed using a combination of topic modeling algorithms and manual thematic coding. The results of this analysis show that government agencies face gaps in technological literacy, unsatisfactory institutional structures, and legal barriers that impede interagency collaboration. We frame these insights from a theoretical perspective on the implementation of digital technology in public administration and provide empirically grounded insights into how these challenges play out in practice. We conclude with a discussion of potential reforms that may be useful to policy and decision-makers in public institutions.Type: journal-articleJournal: Swiss Yearbook of Administrative SciencesVolume: 14Issue: 1DOI: 10.5334/ssas.177