AI Technology: Natural Language Processing | Generative Language Modeling | Deep Learning
This project aims to develop novel methods for adapting language content across cultures, framing cultural adaptation—a key concept from translation studies—as a central task in AI.
The project addresses the challenges of enabling large language models (LLMs) to automatically generate content that is culturally relevant while preserving the original meaning. By compiling multilingual datasets, integrating cultural knowledge, and developing evaluation frameworks, this project will advance AI systems capable of cross-cultural communication. The focus is on overcoming technical barriers such as bias, data sparsity in low-resource languages, and the complexity of evaluating culturally appropriate content.
Daniel Hershcovich, University of Copenhagen, Denmark