Narrative Studio: Visual narrative exploration using LLMs and Monte Carlo Tree Search

2025-04-03

AI Summary

Existing interactive narrative tools are constrained by linear dialogue flows, limiting support for multi-branch “hypothetical” scenario exploration and thereby hindering narrative diversity and depth. This paper proposes a browser-based, tree-structured interactive narrative exploration environment. Our method introduces a novel integration of Monte Carlo Tree Search (MCTS) for automated path selection, augmented by dynamic entity-relation graph modeling to enhance narrative coherence, and real-time frontend tree visualization to enable human-AI collaborative planning. The framework unifies iterative large language model generation, MCTS-based path evaluation, entity-graph-driven consistency constraints, and visual interaction. User studies demonstrate a 3.2× improvement in branch exploration efficiency and a 41% increase in story logical coherence scores, significantly overcoming the limitations of linear narrative paradigms.

Technology Category

• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Planning, Routing, and Scheduling: Planning with Language Models
• Multiagent Systems: Multiagent Planning

Application Category

• Search and Retrieval-Augmented AI: Assisted, interactive, and conversational search
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Interactive storytelling benefits from planning and exploring multiple 'what if' scenarios. Modern LLMs are useful tools for ideation and exploration, but current chat-based user interfaces restrict users to a single linear flow. To address this limitation, we propose Narrative Studio -- a novel in-browser narrative exploration environment featuring a tree-like interface that allows branching exploration from user-defined points in a story. Each branch is extended via iterative LLM inference guided by system and user-defined prompts. Additionally, we employ Monte Carlo Tree Search (MCTS) to automatically expand promising narrative paths based on user-specified criteria, enabling more diverse and robust story development. We also allow users to enhance narrative coherence by grounding the generated text in an entity graph that represents the actors and environment of the story.

Problem

• Enables branching exploration of interactive storytelling scenarios
• Overcomes linear flow limitations in chat-based LLM interfaces
• Automates narrative path expansion using MCTS and entity graphs

Innovation

• Tree-like interface for branching story exploration
• LLM inference guided by user and system prompts
• Monte Carlo Tree Search for automatic path expansion