David Abadi is a postdoctoral research fellow at the Department of Social Psychology, University of Amsterdam, Amsterdam Interdisciplinary Centre for Emotion (AICE) and a member of the H2020-funded project DEMOS (Democratic Efficacy and the Varieties of Populism in Europe). His research examines the psychological and emotional mechanisms linked to populist and extremist sentiments (and their corresponding appraisals) as well as hate speech and abusive language across media. For this purpose, he deploys a combination of (computational) methods, such as big data analytics, machine learning algorithms, cross-national survey research, experimental design, and natural language processing (NLP). He has co-developed a deep learning NLP model based on RoBERTa (Robustly optimized BERT approach), in order to detect metaphor, emotion and political rhetoric in big data. Lately, he co-developed a multi-task learning (MTL) model to detect social identity, populist attitudes, news bias and emotion in social media data.
Specializations
| Methodological | Cross-national Surveys, Psychological Experiments, Textual Analysis, Computational Modeling (NLP) |
| Topical | Psychological predictors of populism. The relationship between populism, emotions, social identity and media bias. |
| Regional | Europe, North America |