Centrifuge: A Visual Tool for Authoring Sifting Patterns for Character-Based Simulationist Story Worlds

Finding characters and events of interest in a large story world with possibly hundreds of characters can be challenging. Story sifting is the method of sorting through a story world’s data to find the desired content. Until now, this process has mainly required someone to use text-based programming or query languages. However, this may prohibit those who prefer a more visual experience when story sifting. We present preliminary work on Centrifuge, a proof-of-concept graphical editing tool that enables users to query a character-based simulated story world for narratively intriguing groupings of characters, character relationships, events, and other entities. Our system presents users with various nodes representing entities in the simulated world or components of the underlying query language’s syntax. Users may author queries by dropping nodes onto the canvas and dragging connections between nodes. Then Centrifuge is responsible for translating their configuration of nodes into a valid query that looks for matching patterns in a simulation data database. Query patterns are meant to be reusable and allow users to hierarchically build new patterns from preexisting ones.