An Imagination-Driven Approach to Computational Creativity for Three-Dimension Scene Generation

<p dir="ltr">This thesis presents an imagination-driven computational framework for systematic creativity generation. The framework enables the structured transformation of minimal user input into diverse creative outputs. It is composed of three distinct layers: Syntax, Semantic, and Pragmatic. The Syntax Layer draws on chemical metaphors and psychological theory, translating abstract intentions into geometric forms and computational parameters through formal rule-based mappings. The Semantic Layer is grounded in Eastern philosophy, embedding cultural meaning and symbolism by mapping geometric forms to I Ching trigrams and dynamically organising scenes using Yao-line inversion and Bagua lattices. The Pragmatic Layer ensures spatial coherence and adaptability by applying biological principles, including allometric scaling and physics-based animation.</p><p dir="ltr">To demonstrate the effectiveness of the framework, a 3D scene generation system has been developed as a proof of concept. This system shows how abstract user intentions can be systematically converted into culturally meaningful and interactively animated 3D visualisations. Each generated scene consists of two symbolic objects and dynamic interactions, preserving the cultural and spatial integrity of the original input.</p><p dir="ltr">The key contributions of this work are as follows. The research establishes imagination as both the source and methodological foundation for creativity. It implements a fully structured three-layer framework that separates and coordinates the roles of syntax, semantics, and pragmatics. The work develops an interdisciplinary mechanism that links psychology, chemistry, philosophy, and biology for a richer model of computational creativity. The framework enables symbolic cultural generation, turning abstract intentions into meaningful 3D symbols. It also achieves a synthesis of Eastern philosophy and Western science to establish a culturally inclusive and scientifically robust approach.</p><p dir="ltr">By bridging human intent and computational processes, this research offers a novel and interpretable methodology for creative systems and lays a foundation for future interdisciplinary applications.</p>