Program

  • 9:00-9:15: Opening Session
  • 9:15-10:30: Elevator Pitches
    • Sara Buchmann: AgentRAG: A Multi-Agent System Combining LLMs with Knowledge Graphs for Trustworthy Auditing paper
    • Theodoros Aivalis: Towards Interpretable Generative AI via Search and Knowledge Graphs paper
    • Antonio Balordi: A Principled Framework for Parameter Estimation in Federated Unlearning paper
    • Verdiana Schena: Semantic Similarity through Human-in-the-loop in Knowledge Graph Embeddings paper
    • Alessia Donata Camarda: Neurosymbolic AI in Digital Forensics: Commonsense and Qualitative Reasoning paper
    • Mario Mirabile: Designing Trustworthy AI Systems for Human-Centric Collaboration paper
    • Diogo Cosme: AI Assistants and Agents in Geographic Information Systems paper
    • Marek Dědič: Representation learning for structured data paper
    • Katharina Kösseler: Automated Annotations enable Large-Scale Morphological Analysis in Kidney Histopathology Using Deep Learning paper
    • Pablo Torrijos Arenas: Federated Learning of Probabilistic Graphical Models paper
    • Nicolò Donati: Generative AI for Innovative Writing and Editing Support Systems paper
    • Frida Hartman: Bias in AI-based recruitment tools paper
    • Estherine Goh: AI and Elderly: Co-Designing Meaningful Human-AI Collaboration and AI Education in Aging Contexts and Intergenerational Communities paper
    • Shanika Edirisinghe: Advanced Techniques for Cloud Obstruction Reconstruction in Satellite Imagery: Exploring GANs, Diffusion Models, and Alternative Approaches paper
    • Alberto Rovetta: Action-Failure Resilient Planning paper
    • Muhan Hou: Active Robot Learning from Demonstrations for Sub-Optimal Human Teaching paper
    • Mohamed Wadhah Mabrouk: Modeling Historical Color Knowledge Leveraging LLMs, Ontologies, and Knowledge Graphs paper
    • Kevin Fee: Advancing Radiogenomics in Prostate Cancer Research via new data science methods paper
    • Alexis McGuire: AI Synthesised Faces: Does Training Improve Detection Performance paper
    • Ejdis Gjinika: Understanding Semantics in Neural Language Models via Learning Trajectories paper
    • Artemis Dampa: Enhancing Scientific Research through Knowledge-Informed AI paper
  • 10:30-11:30: Poster session + Coffee Break
  • 11:30-12:30: Invited talk
    • Giuseppe Marra, KU Leuven, Belgium
    • Title: Towards a novel perspective of neurosymbolic AI
    • Abstract: This talk presents an overview of Neurosymbolic AI, a field that seeks to integrate deep learning with reasoning over symbolic knowledge. To systematically explore this integration, the presentation adopts dimensions and concepts from statistical relational learning. A blueprint will be introduced for both current and prospective neurosymbolic systems, demonstrating how standard formal systems can be integrated with neural networks by leveraging probabilistic models as interfaces. After presenting several concrete instantiations of this blueprint, the talk will conclude by advocating for a fundamental reconceptualization of Neurosymbolic AI, shifting its focus from a problem of knowledge integration to one of human-machine specification.
    • Bio: Giuseppe Marra is Assistant Professor at the Department of Computer Science at KU Leuven since 2022, where he coleads the Neurosymbolic lab, together with Prof. Luc De Raedt. His research interests are in theories and frameworks of neurosymbolic AI, with a particular focus on logical and probabilistic reasoning, as well as applying neurosymbolic techniques to relational and interpretable deep learning models. He has contributed to several pioneering Neurosymbolic frameworks, including Semantic Based Regularization, DeepProbLog and DeepStochLog. He is the author of an authoritative survey on the subject and serves as the principal investigator for multiple national and European projects on neurosymbolic AI and its applications.
  • 12:30-13:30: Lunch
  • 13:30-14:15: Plenary Discussion
  • 14:15-15:30: Elevator Pitches
    • Emna Lakani: A Neuro-symbolic knowledge graph framework for geological data integration and interoperability paper
    • Michele Dusi: Following Stereotypes in LLMs’ Learning Trajectories paper
    • Christodoulos Kechris: The unreasonable effectiveness of dissecting models: A case for trustworthy and efficient biomedical time-series models paper
    • Roberto Barile: Neural-Semantic Methods for Knowledge Graph Completion and Explanation paper
    • Vojtěch Bláha: Efficient Black-box Optimization via Deep Surrogate Models and Transfer Learning paper
    • Nicolo Mombelli: Interpretable Language Models for Transparent Decision-Making in Business-Critical Domains paper
    • Sarra Djebour: Richer Oracle Interactions in Constraint Acquisition: Theory and Application to Program Verification paper
    • Ainhoa Vivel Couso: Empirical Study on the Energy Efficiency of Transfer Learning Techniques for Text-to-Text Generation paper
    • Aicha Boukhari: Toward Practical Constraint Acquisition: An Anytime Learning Approach paper
    • Davide Baldini: Human oversight in Automated Decision-Making: A Legal Safeguard Against Algorithmic Discrimination Under EU Law paper
    • Patrycja Kwiek: Color Matters: Evaluating the Role of Color Space Modeling and Color Analysis in Biomedical Image Interpretation Using Machine Learning paper
    • Valerio Borelli: Beyond Graph Neural Networks Expressivity: Topological Learning for Classical Planning paper
    • Grazia Ferrara: Designing Deployable Public Health Campaigns via Online Learning Techniques paper
    • Elena Yan: Self-Adaptive Regulation Mechanisms for a Trustworthy and Sustainable Industry of the Future paper
    • Anica Cvetkovic: Interactions With Smart Machines: Can We Trust AI? paper
    • Victor Lavairye: Modeling and simulation of human behavior for human-centric digital twins paper
    • Grafika Jati: False Confidence: Is Detection Reliable Enough for Autonomous Vehicles in LiDAR Point Cloud Corruption? paper
    • Aysu Bogatarkan: Flexible, Lifelong, Explainable, and Robust Solutions for Multi-Agent Path Finding Problems paper
    • Fabio Michele Russo: Perceptions of Explainable AI: how presentation is content paper
    • Gianmarco Pappacoda: Entity Name Recognition: a LLM-friendly set-based approach to Knowledge Extraction paper
    • Rishabh Shukla: Biometrics in the Age of Generative AI paper
    • Juan Camilo Rosero Lopez: Explainable Multi-Objective Reinforcement Learning paper
  • 15:30-16:30: Poster session + Coffee Break
  • 16:30-17:30: Career Panel
  • 17:30-17:45: Closing Session