Presentation Schedule
Local Memory Mosaic: Narratives of Local Gazetteers Under the Generative AI-Driven Paradigm (91450)
Monday, 12 May 2025 16:30
Session: Conference Poster Session
Room: Orion Hall (5F)
Presentation Type: Poster Presentation
Local gazetteers are known as local encyclopedias, recording in detail important local events from ancient times to the present. Due to their extensive length, typically spanning hundreds of thousands of words, events are often dispersed across chapters, posing challenges for readers seeking comprehensive understanding. This study, under the AI for Science (AI4S) paradigm, explores how generative AI, particularly LLMs, can interpret the humanistic dimensions of gazetteers and generate coherent narrative discourse. Drawing on Gérard Genette’s narrative theory, the study identifies three narrative levels: (1) the story layer, which represents the core event conveyed by the discourse content; at this level, generative AI primarily relies on its language comprehension capabilities to extract narrative elements and logical relationships from local gazetteers, constructing a coherent storyline; (2) the native narrative layer, which refers to the oral or written discourse describing the narrated event. Using generative AI’s content generation capabilities, story texts, images, and even video clips are produced based on the available materials, transforming monotonous historical facts into engaging humanistic stories with dynamic plots; and (3) the narrating layer, which pertains to the process of storytelling itself. Through the human-computer interaction capabilities of generative AI, this level enhances the appeal of the story by incorporating anthropomorphic language styles, user-friendly interactive designs, and immersive narrative scenes. Finally, this study takes Zhejiang Tongzhi as the local gazetteer data to develop a knowledge base platform for West Lake stories in China, comprising modules such as in-depth gazetteer reading, cultural perspective analysis, and intelligent story generation.
Authors:
Anrunze Li, Renmin University of China, China
Li Niu, Renmin University of China, China
Chi Jin, Renmin University of China, China
Canbin Chen, Renmin University of China, China
Lei Wang, Sun Yat-Sen University, China
About the Presenter(s)
Anrunze LI is a second-year PhD student in Information Resources Management at School of Information Resource Management, Renmin University of China.
See this presentation on the full schedule – Monday Schedule
Comments
Powered by WP LinkPress