Ashrakat Elshehawy
noon - 1 pm
Large Language Models for Social Science at Scale
Ashrakat Elshehawy
The session will show how social scientists can use LLMs to extract structured information from messy, unstandardized textual data, such as media reports, websites, or administrative documents, at scale. We will cover practical prompt-engineering techniques. The session will walk through simple, adaptable code for classification and information extraction, and introduce basic verification strategies (performance metrics, human-in-the-loop checks, and consistency tests) to ensure reliability.
Nikolay Marinov
10 am - 1 pm
Intro Py: Workshop Part I
Nikolay Marinov
This six-hour workshop, divided into two sessions, is a practical introduction to Python for people with no prior experience. Participants will be able to interpret, modify, and create scripts, run models, save output - and will have a foundation for further learning.
Plan:
Github: a general introduction to data communuty’s go to place for storing and sharing code
Where to run Python
Write, run, and troubleshoot simple code
Variables, data types
Control structures, conditionals and loops
Lists, dictionaries, user input
Functions, modular code
Loading data and saving data
Py extensions: for “statistics”, language analysis, visuals
Py: how polisci currently uses it for research and automation
10 am - 1 pm
Intro Py: Workshop Part II
Nikolay Marinov
Dec 5 2025 - Erica Metheney and Lauren Yehle
Nov 21 2025 - Ann-Ida Scheiber Gyllenspetz
Nov 3 2025: Introduction to visualisation principles by Jonathan Westin and Matteo Tomasini
Recommended readings:
Edward R. Tufte 1969. “Improving Data Analysis in Political Science”. World Politics, Vol. 21, No. 4 (Jul., 1969), pp. 641-654 (14 pages) https://doi.org/10.2307/2009670 (https://www.jstor.org/stable/2009670)
Vertesi, J. Theory-Laden Data Visualization, Drawing-As and Seeing-As in Sociology and in Data Science. Am Soc 56, 440–454 (2025). https://doi.org/10.1007/s12108-025-09658-2
Nov 4 2025: From unstructured data to visualisations, lead by David Alfter
Recommended readings:
Chs 2 and 17 in Jurafsky, D., & Martin, J. H. “Speech and language processing: An introduction to natural language processing, computational linguistics, and speech recognition with language models (3rd ed., draft). Stanford University; University of Colorado at Boulder. https://web.stanford.edu/~jurafsky/slp3/ Wednesday, Nov 5 (13-16)
On Nov 5 2025, we can hear more on Visual data: from raw data to publication by Matteo Tomasini
On Nov 6 2025, Matteo Tomasini talks about “Bases of spatial visualisation”
Oct 17 2026 Gard Olav Dietrichson:
What You’ll Learn: Structural Topic Modeling in R
This session introduces how to uncover hidden themes in large text collections using Structural Topic Modeling (STM). Participants will learn to:
Prepare and clean text data for analysis.
Tokenize and filter words, removing stopwords and applying word stemming.
Use TF-IDF to identify meaningful words and visualize word frequency patterns.
Build and interpret an STM, discovering key topics and how they vary across groups (e.g., by gender).
Load and prepare labeled text data, cleaning and restructuring it for analysis.
Convert model predictions into binary classifications for comparison.
Join automated and manual classifications to assess accuracy.
Use confusion matrices to measure model performance across different topics.
Oct 10 2025 - Nikolay Marinov - Machine-Learning to Classify Images:
Beginner-friendly workshop introducing how machine learning can be used to analyze and classify images — from detecting faces and identifying objects to training simple models that distinguish between categories such as male and female politicians’ social media banner images. Using Google Colab and Python libraries like OpenCV, Hugging Face’s transformers, and PyTorch, participants learn the full workflow: setting up datasets, applying pre-trained models for visual recognition, labeling and training supervised models, and evaluating performance metrics like accuracy and F1 scores. We use the social media banner pictures of members of the European Parliament.