BookWorld：從小說到互動代理社會，用於創意故事生成

BookWorld: From Novels to Interactive Agent Societies for Creative Story Generation

April 20, 2025

作者: Yiting Ran, Xintao Wang, Tian Qiu, Jiaqing Liang, Yanghua Xiao, Deqing Yang

cs.AI

摘要

大型語言模型（LLMs）的最新進展使得通過多智能體系統進行社會模擬成為可能。先前的研究主要集中在從零開始創建的智能體社會，為智能體分配新定義的角色。然而，模擬已建立的虛構世界和角色在很大程度上仍未得到充分探索，儘管其具有重要的實用價值。在本文中，我們介紹了BookWorld，這是一個用於構建和模擬基於書籍的多智能體社會的綜合系統。BookWorld的設計涵蓋了全面的現實世界複雜性，包括多樣化且動態的角色、虛構的世界觀、地理限制及其變化等。BookWorld支持多種應用，包括故事生成、互動遊戲和社會模擬，提供了擴展和探索喜愛虛構作品的新方式。通過大量實驗，我們證明BookWorld能夠生成創意且高質量的故事，同時保持對原著的高度忠實，以75.36%的勝率超越了先前的方法。本文的代碼可在項目頁面找到：https://bookworld2025.github.io/。

English

Recent advances in large language models (LLMs) have enabled social simulation through multi-agent systems. Prior efforts focus on agent societies created from scratch, assigning agents with newly defined personas. However, simulating established fictional worlds and characters remain largely underexplored, despite its significant practical value. In this paper, we introduce BookWorld, a comprehensive system for constructing and simulating book-based multi-agent societies. BookWorld's design covers comprehensive real-world intricacies, including diverse and dynamic characters, fictional worldviews, geographical constraints and changes, e.t.c. BookWorld enables diverse applications including story generation, interactive games and social simulation, offering novel ways to extend and explore beloved fictional works. Through extensive experiments, we demonstrate that BookWorld generates creative, high-quality stories while maintaining fidelity to the source books, surpassing previous methods with a win rate of 75.36%. The code of this paper can be found at the project page: https://bookworld2025.github.io/.