The “Machine Learning for Adaptive Teaching (MATE)” project systematically investigates adaptive teaching and learning supported by generative AI, aiming to translate evidence-based findings into sustainable school practice. The project seeks to develop AI-powered learning systems that respond to individual student needs—thereby fostering both lasting learning outcomes and equity in education.
MATE builds on prior work on technology-enhanced adaptive instruction, including AI-based tutoring systems and task generators, and consolidates these efforts into a coherent long-term research program. The project pursues three strategic goals: advancing transdisciplinarity by integrating educational research, psychology, subject-matter didactics, ethics, and AI; building sustainable AI infrastructure for data science and adaptive interventions; and ensuring long-term impact through pre-service and in-service teacher education. Research is organized into four thematic squads covering the psychological foundations of adaptive AI systems, co-constructive design and implementation, data science and effectiveness research, and teacher education and public engagement.
MATE is led by Prof. Dr. Andreas Lachner at the Tübingen Center for Digital Education (TüCeDE) and embedded in the Tübingen ecosystem, with close collaborations involving the ELLIS Institute Tübingen, the Tübingen AI Center, the LEAD Graduate School and Research Network, and the Centre for School Quality and Teacher Education (ZSL). International partnerships connect the project with the National Education Lab AI (NOLAI, Netherlands), the Centre for Learning Analytics at Monash University (CoLAM, Australia), and the Human Computer Interaction Institute at Carnegie Mellon University (CMU, USA).
The project is funded by the Volkswagen Foundation under the “Momentum” programme with a total grant of €849,600 (direct project costs) over a period of 4 years.
Funding acknowledgement: This project is funded by the Volkswagen Foundation under the “Momentum” funding programme.
