Key points are not available for this paper at this time.
We argue that forcing users to describe visual aesthetics with text represents a fundamental architectural flaw in digital fashion systems. We propose a dual-layer framework as a conceptual alternative that reframes style discovery from a linguistic task to a spatial navigation experience in eXtended Reality (XR). The system's upper layer maps fashion styles into a navigable 3D manifold based on 15 affective dimensions derived from Kansei Engineering. The lower layer contains 3,000 designer-curated outfit configurations with detailed structural annotations. Users explore the style landscape through spatial movement and refine their selections through a Dual-Layer Style-Structural Mapping interface radar, which maps real-time affective adjustments to corresponding garment structures. We contend that this spatialized approach bypasses semantic barriers and provides designers with objective behavioural data through recorded navigation trajectories. We discuss design details, interaction methods, and key challenges, including cultural bias, interpretability, and scalability.
Li et al. (2026) studied this question.