The SCOPE project’s CSSM workshop series continued its training agenda with a three-day immersive workshop on Natural Language Processing (NLP) from August 6-8, 2025. Led by instructor Indira Sen (U. Mannheim, GESIS), the event gave 10 participants a comprehensive overview of NLP methods and the opportunity to apply newly learned techniques to their own research projects.
The workshop opened with the core building blocks of the discipline, exploring language modeling, text classification, and the concept of vector semantics and embeddings. Sessions then progressed to more advanced techniques, covering contextual embeddings, the mechanics of attention, and the role of transformer-based models like BERT.
A key part of the training also focused on the practical steps of data cleaning and its effects on text analysis results.The final day was dedicated to application and practice. Participants learned how to fine-tune pre-trained models for specific tasks and experimented with using large language models for content analysis.
The workshop concluded with a hands-on session where attendees could work directly on their own projects with guidance.By the end of the three days, participants had gained both a conceptual and practical understanding of NLP. The combination of theory, demonstrations, and personal experimentation ensured they left equipped with tangible skills to apply to future projects.