Proposals from M. Brambilla

• Explainable AI (XAI)
  • Explainability of CNNs
  • Explainability of tabular data
  • Concept-based Explainability
  • Human-in-the-loop Explainability
  • Gamified Data Collection for NLP Explainability Tasks
  • Gamified Approaches to Evaluate the Understandability of Explainations
• Large Language Models (LLMs), Neuro-Simbolic AI, and Kwnoledge Management
  • Augmenting LLMs with domain-specific knowledge with Retrieval Augmented Generation (RAG)
  • Coupling RAG with Knowledge Graphs (KGs) to augment LLMs
  • Analyze the use, configuration, and features of LLMs
  • Implement and analyze LLMs as generator of UML/IFML or BPMN models
• Data Science
  • Combining Big Data, Data Engineering, and Data Science (ML, dataviz, NLP) for:
    • Societal challenges
    • Socio-political settings
    • Economy, ethics
    • News and disinformation
• Extraction and analysis of medical images from Electronic Medical Records (EMRs)