Syllabus: GESIS Fall Seminar course Advanced Text Analysis | Hauke Licht

Super excited to teach "From Embeddings to LLMs: Advanced Text Analysis with Python" at the GESIS Training Fall Seminar of Computational Social Science this week 🥳 We will - Review the role of conceptualisation and manual annotation in computational text analysis. Practice to form concepts and annotate their occurrences in text materials. - Learn about static and contextualised word embeddings. - Work with transformer encoder models for (inductive) topic modelling and (deductive/supervised) text classification through transfer learning (a.k.a. fine-tuning). - Work with large pre-trained generative transformer (GPT) models (a.k.a LLMs) to classify texts through in-context learning (a.k.a. prompting). - Practice and explore advanced techniques and applications. Discuss research ideas and viable measurement strategies in 1-on-1s with the instructor. #TextAsData #NLP #teaching #Python #LLMs #GESISfallseminar

tubecleanR – GESIS - KODAQS Toolbox

Forschungsnah und innovativ: Rückblick auf die VDB-Fortbildung für Fachreferent*innen der Soziologie und Politikwissenschaft | o-bib. Das offene Bibliotheksjournal / Herausgeber VDB

EconStor: A Replication Report on "Political polarization of news media and influencers on Twitter in the 2016 and 2020 US presidential elections" by Flamino et al. 2023

EconStor is a publication server for scholarly economic literature, provided as a non-commercial public service by the ZBW.

Reproducibility & Replicability in Computational Social Science Research

Presentation on "Reproducibility & Replicability in Computational Social Science Research" given at the workshop "Computational Social Science: AI and Society – Exploring Inequality in the Digital Age", Mannheim, Germany, May 15-16, 2025.