Title: Bridging the Event-Character Divide in Computational Narrative Understanding

Abstract: As AI systems increasingly function as a social and cultural technology, their ability to comprehend narratives and stories -- the primary way humans create shared meaning -- has become a critical area of research. However, computational approaches to this problem are often fractured, with some models focusing on the sequence of events and others on the psychology of characters. This talk brings a unified perspective to this landscape by operationalizing a narrativity framework that synthesizes insights from multiple disciplines. Using this framework as our guide, we will survey computational methods for narrative understanding, present a case study of a unified model, and demonstrate how this integrated view enables the discovery of powerful narrative themes that shape public opinion on critical social issues.