Time: 11:30 AM
Location: CB01.27.014 & Teams
Neurosymbolic AI for Education
Eleni Ilkou • Leibniz Information Centre for Science and Technology, Hannover, Germany
CB:01.27.014 & Teams
Abstract: Neurosymbolic AI invites us to reconsider how knowledge is represented, reasoned with, and communicated in educational systems. The symbolic dimension, embodied in Knowledge Graphs and ontologies, enables reasoning, and explainability. When combined with neural and statistical methods, such as Language Models, these representations give rise to interactions that are transparent, semantically grounded, and pedagogically meaningful. This talk draws on recent works to explore how these two layers, the symbolic and the neural, can be aligned to support interpretable, adaptive, and pedagogically coherent systems. By linking structured semantic representations with data-driven insights, Neurosymbolic AI can foster learning environments that support conceptual understanding and contextual adaptation, opening pathways for adaptive content delivery and advanced learning analytics across diverse educational settings, from self-regulated and collaborative e-learning to the generation of accessible learning materials.
Bio: Eleni Ilkou is a researcher in AI and education affiliated at the TIB – Leibniz Information Centre for Science and Technology in Hannover, Germany. She is currently in Australia on a Visitor Scholarship by the ARC Training Centre for Information Resilience at the University of Queensland, working with Prof. Gianluca Demartini. Her research lies at the intersection of AI, Semantic Web, Knowledge Graphs, and Language Models, with applications in collaborative and self-regulated e-learning, and smart learning environments. Eleni’s work has appeared in top-tier venues such as IEEE Transactions on Learning Technologies (TLT), The Web Conference (WWW), International Semantic Web Conference (ISWC), and Artificial Intelligence in Education (AIED), earning a Best Paper nomination at AIED 2025 and an Honourable Mention at ISWC 2024. [Google Scholar]