A Knowledge Distribution Model to Support an Author in Narrative Creation

Abstract

Adjusting the knowledge of characters and the reader is a critical task for an author in narrative creation. Throughout a narrative, both characters and the reader experience events according to their own timelines and perspectives. They interpret information accumulated through their experience and update knowledge to the narrative-world which the author constructed. In this paper, we present a Knowledge Distribution Model which supports an author in finely controlling the knowledge of characters and the reader. Within the model, the Knowledge Structure is constructed by connecting event, information, and knowledge. The Knowledge State is evaluated as the degree of belief under the knowledge structure. We adopted a probabilistic reasoning model to calculate the knowledge state. The change in knowledge state, defined as Knowledge Flow, is visually presented to the author. We designed a GUI prototype to implement the proposed modeling process, and demonstrated the knowledge flow with an actual cinematic narrative.