Evaluating Gender Bias in German Machine Translation (6 LP) - Let’s explore the Future of Fair AI together!

Machine translation systems (MT) have become an indispensable part of our daily lives – they translate texts, subtitles, and even spoken language in real time. While they help us overcome language barriers, critical questions arise: Do these systems perpetuate harmful stereotypes? Do they reinforce biases instead of reducing them? And consequently: How can we even measure this? In this interdisciplinary research project, we are tackling these very questions!

What's it about?

Our goal is to develop an improved Gender Bias Evaluation Testset (GBET) that systematically evaluates gender bias in machine translations. We build upon an existing method and expand it through new data and approaches. We examine various translation models - from Google Translate to DeepL to large language models like ChatGPT - and evaluate the extent to which they exhibit gender bias in different ways.

What will you learn and do?

• Critical analysis of machine translation
• Development of a testset for systematic bias measurement
• Optimization of an automatic evaluation pipeline using Python
• Work with popular machine translation models
• Scientific work with the opportunity to contribute to a publication and present this work at a conference

Who is this module suitable for?

The project, funded by the Berlin University Alliance, is aimed at students from the fields of Computer Science, Linguistics, Social Sciences, or related disciplines. You don't need extensive prior knowledge in machine translation, but interest in language, AI, and fairness is important. Knowledge of Python is helpful but not required.

Important: Students from all Berlin universities (TU, FU, HU, Charité) can participate and get ECTS - no need to apply for guest membership!

Why should you join?

This module offers you a unique opportunity to work on current research, deepen your knowledge in AI and language technology, and make a real contribution to fairness in machine translation. Your work could help make future translation systems more equitable!

Join ISIS course now: https://isis.tu-berlin.de/course/view.php?id=42828 (Note: The link is initially only accessible to TU students, all other students will get access later!)