|10 am - 5 pm (Friday 10 am - 2.30 pm)|
Sonderforschungsbereich 1342 "Globale Entwicklungsdynamiken von Sozialpolitik", Universität Bremen
Day 1 - Sept. 30th - 10am to 5pm (roughly) - Overview of Text Analysis
The first day will combine a comprehensive overview of quantitative text analysis with an introduction to the structure, logic, and syntax of the quanteda text analysis package. Topics will include:
- an introduction to quantitative text analysis and its workflow
- a comprehensive map of quantitative text analysis including the main techniques used in political and social science
- an overview of quanteda and how to use it to create core objects for textual analysis.
Day 2 - Oct. 1st - 10am to 5pm (roughly) - Describing, comparing, and scaling texts
The second day will use a few running examples including:
- identifying key words based on statistical association measures
- identifying multi-word expressions via collocation analysis
- computing and interpreting textual statistics for comparing texts
- using bootstrapping to measure uncertainty in texts
- computing similarities and distances for identifying clusters of similar texts.
Day 3 - Oct. 2nd - 10am to 2:30pm (must end before Colonialism workshop) - Advanced text analysis
The final day will cover advanced methods including:
- feature weighting and feature selection for textual matrices;
- supervised and unsupervised document scaling for measuring ideological positions;
- supervised machine learning for document classification; and
- topic modelling.