computational social science
Definition
Computational social science is an interdisciplinary approach that applies automated data collection and machine learning methods to analyze human behavior at scale. In the study of political communication, this means extracting quantitative behavioral data from large video corpora, such as facial expression recognition scores drawn from 220 YouTube videos of global political leaders, using transformer-based zero-shot learning models. The resulting time-series data can then be submitted to network psychometric analyses, here dynamic Exploratory Graph Analysis, to map the structural organization of emotional co-expression across both populist and non-populist leaders. Such an approach treats non-verbal signals as analyzable data streams rather than qualitative impressions, enabling systematic comparisons that would be intractable through manual coding.
Sources: Tomašević & Major (2024)
Related Terms
- network analysis (1 shared article)
- emotions (1 shared article)
- exploratory graph analysis (1 shared article)
- populism (1 shared article)
- affective dynamics (1 shared article)
Applications
Computational Social Science and Facial Expression Recognition
Facial expression recognition serves as a primary data-generation mechanism within computational social science research on political communication. Automated detection of six emotions, happiness, excitement, hope, anger, fear, and sadness, from video footage allows researchers to construct emotion time series without relying on human coders, reducing subjectivity and enabling analysis at a scale that manual methods cannot match.
Sources: Tomašević & Major (2024)
Computational Social Science and Affective Political Communication
Computational methods offer a means of examining emotional dynamics in political speeches that qualitative and survey-based approaches leave largely unaddressed. By combining machine learning-based emotion detection with network psychometric modeling, researchers can identify how discrete emotions covary over time and how that covariance structure differs between leaders who deploy varying degrees of populist rhetoric.
Sources: Tomašević & Major (2024)
