Little Red Riding Hood Goes Around the Globe:Crosslingual Story Planning and Generation with Large Language Models
This work addresses the problem of generating coherent stories in multiple languages for applications in creative writing or language learning, but it is incremental as it extends monolingual planning methods to a cross-lingual setting.
The study tackled cross-lingual story generation by investigating whether planning improves narrative quality across languages, finding that three-act plans led to more coherent and interesting stories while enabling explicit control over content and structure.
Previous work has demonstrated the effectiveness of planning for story generation exclusively in a monolingual setting focusing primarily on English. We consider whether planning brings advantages to automatic story generation across languages. We propose a new task of cross-lingual story generation with planning and present a new dataset for this task. We conduct a comprehensive study of different plans and generate stories in several languages, by leveraging the creative and reasoning capabilities of large pre-trained language models. Our results demonstrate that plans which structure stories into three acts lead to more coherent and interesting narratives, while allowing to explicitly control their content and structure.