Massimo Stella is Associate Professor of AI and Psychometrics at the University of Trento. As PI of CogNosco Lab, he leads PENSO, a research project investigating how LLMs can influence human cognition. His research advances interpretable AI, cognitive network science, and computational approaches to knowledge modelling.

Lecture at Complexity Meets Heritage:
Reading Between the Nodes: Interpretable NLP and Semantic Framing for Text Analysis

How can we make computational interpretations of novels transparent, reproducible, and open to criticism? This hands-on tutorial introduces an interpretable complex systems pipeline for extracting insights from cultural texts. Participants will transform passages into textual forma mentis networks and target-event-agent networks, i.e. knowledge graphs linking concepts through syntactic, semantic and emotional associations. We will examine central concepts, communities, and local semantic frames to investigate how characters, themes, and emotions are portrayed. Through practical but interactive coding examples, we will compare narrative perspectives, visualize competing frames, and test robustness against preprocessing choices and appropriate null models. No prior network science experience or Python familiarity is required, as we will have the assistance of coding AIs.