Narrative-Driven Paper-to-Slide Generation via ArcDeck

2026-04-13

AI Summary

Existing approaches to converting academic papers into slide presentations often neglect the original text’s logical structure and narrative coherence. This work reframes the task as a structured narrative reconstruction problem and introduces, for the first time, explicit discourse structure modeling combined with a multi-agent collaborative mechanism. By constructing discourse trees and maintaining a global commitment document, the method preserves high-level authorial intent, while a role-specialized multi-agent critique-and-revision framework iteratively refines both the presentation outline and visual layout. Evaluated on ArcBench—a newly curated benchmark—the proposed approach significantly outperforms current state-of-the-art methods in terms of narrative fluency and logical consistency.

Technology Category

• Natural Language Processing: Discourse, Pragmatics & Argument Mining
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Multiagent Systems: Adversarial Agents

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

We introduce ArcDeck, a multi-agent framework that formulates paper-to-slide generation as a structured narrative reconstruction task. Unlike existing methods that directly summarize raw text into slides, ArcDeck explicitly models the source paper's logical flow. It first parses the input to construct a discourse tree and establish a global commitment document, ensuring the high-level intent is preserved. These structural priors then guide an iterative multi-agent refinement process, where specialized agents iteratively critique and revise the presentation outline before rendering the final visual layouts and designs. To evaluate our approach, we also introduce ArcBench, a newly curated benchmark of academic paper-slide pairs. Experimental results demonstrate that explicit discourse modeling, combined with role-specific agent coordination, significantly improves the narrative flow and logical coherence of the generated presentations.

Problem

• paper-to-slide generation
• narrative coherence
• logical flow
• discourse structure
• presentation generation

Innovation

• narrative-driven generation
• discourse tree
• multi-agent framework
• paper-to-slide conversion
• structured narrative reconstruction