Story Ribbons: Reimagining Storyline Visualizations with Large Language Models

Area of Science:

• Digital Humanities
• Computational Linguistics
• Literary Studies

Background:

• Analyzing literary texts involves complex relationship mapping (characters, themes).
• Extracting structured data from unstructured narratives is a significant challenge.
• Advancements in Large Language Models (LLMs) offer new possibilities for text analysis.

Purpose of the Study:

• To leverage LLMs for automated narrative information extraction.
• To develop an interactive visualization system for literary analysis.
• To explore LLM capabilities in augmenting storyline visualization techniques.

Main Methods:

• Developed an LLM-driven data parsing pipeline for novels and scripts.
• Created STORY RIBBONS, an interactive visualization system.
• Conducted evaluations and user studies on 36 literary works.

Main Results:

• Demonstrated LLMs' potential to streamline narrative visualization creation.
• Showcased STORY RIBBONS' ability to reveal insights into character and theme trajectories.
• Identified current limitations of AI-based systems in literary analysis.

Conclusions:

• LLMs can significantly enhance the process of literary visualization.
• Interactive systems like STORY RIBBONS facilitate deeper exploration of narrative structures.
• Further research is needed to address AI limitations and refine interaction motifs.