Research Summary: Studying Comics at Scale with Artificial Intelligence: The PAGE Framework and Three Studies of American Comics and Japanese Manga

Researchers have introduced the PAGE framework, an artificial intelligence methodology designed to scale comics research across extensive historical corpora, specifically examining American comics and Japanese manga. The framework addresses limitations in traditional scholarship—which has historically analyzed works on an individual basis—by combining automated segmentation of panels, text, and metadata with vision-language model annotations governed by codebooks and grounded in human verification.

Why it matters

This methodology indicates a viable pathway for unlocking large-scale visual and cultural archives that were previously intractable using manual analysis alone. It demonstrates how modern vision-language models can be systematically aligned with formal qualitative research codebooks. For institutions managing vast visual collections, it provides an operational model to extract historical and thematic trends at scale.

Key insights

• Comics constitute a vast twentieth-century visual record documenting societal depictions of leaders, wars, and heroes, but scale constraints have historically limited scholarship to single-book analyses.
• The proposed PAGE framework utilizes AI and generative models to parse scanned pages systematically into panels, textual components, and bibliographic metadata.
• The framework uses vision-language models following structured, written codebooks to replicate human coding methods at scale.
• Computational readings within the framework are grounded and validated through human annotation samples to explain cross-corpus changes.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2610.13840