We're drowning in textual data—especially in fields like biomedicine where COVID-19 alone generated massive public datasets. How do we automatically extract meaningful knowledge from millions of research papers and technical documents? This seminar presents a comprehensive approach to knowledge management that overcomes the traditional semantic barriers of unstructured text processing. The research explores two complementary methodological frameworks: bottom-up approaches that discover latent document structures, and top-down methods that organize knowledge into structured representations like knowledge graphs. Combined together, these techniques unlock rapid knowledge discovery and empower users (both humans and autonomous agents) to navigate complex domains with minimum effort.
Short Bio: Francesco Invernici is a PhD candidate in Information Technology at Politecnico di Milano. His research areas are knowledge management, graph databases and natural language processing, with a strong interest for unsupervised learning. In 2023, his project "Topics Evolution That You See" has been selected for funding by the NGI Search initiative.
