AIGC-assisted geospatial generative design for classical Chinese literary works

Visualizing the geographic space in the text is a good way to reveal the spatial structure of the narrative. The spatial maps in current fictional works require the creator to carefully read and analyze the entire text of the literary work and then draw it manually, which consumes a lot of time and manpower.This study proposes a systematic method that combines natural language processing (NLP) and artificial intelligence generated content (AIGC) to transform the geographic space described in classical texts into three-dimensional (3D) visualization maps. The study first designs a three-stage NLP method of “positioning-divided-judgment” to accurately extract geographic entities in the text; secondly, formulates and applies rules for converting literary descriptions into quantified spatial coordinates; finally, the 3D-AIGC model is used to combine descriptive poems to generate geographic entity models and arrange them into maps according to their spatial coordinates. The text-geospace rapid conversion method proposed in this study can assist the author’s literary creation and the reader’s reading comprehension, and provides new ideas in the fields of text digital display and film and television adaptation, showing the application potential of AI in the fields of literature and digital humanities.