Atomically True Ontology Modelling: Residential Buildings

Abstract

AbstractCreation of realistic residential buildings has been the topic of research in procedural content generation (PCG) for several years. Most PCG methods rely on analysing domain knowledge in the applicable setting and encoding that knowledge as configuration for the PCG methods used. While this technique does solve the immediate problem, it is often not portable, and extensions of a specific generation technique often requires reconfiguring when new or more detailed domain knowledge is attained. This paper presents a holistic ontological driven PCG technique that allows for varied PCG methods to be used and interchanged while having a stable platform for domain knowledge. The technique presented in the paper is an implementation of a PCG framework: atomically true ontology modelling (ATOM). Using ATOM for creating residential buildings allows for the definition of general concepts such as rooms and room types that can be then used to drive PCG methods to produce residential buildings in distinctive styles and environments.KeywordsProcedural content generationBuilding plansOntology developmentGenerative design