Evidence-Weighted Heterogeneous Graph Mapping for Symmetry–Asymmetry Analysis in AI-Assisted Electric Motorcycle Morphology Design

Electric motorcycle morphology design involves relationships among whole-vehicle profiles, exposed local modules, affective semantics, evaluation criteria, and AI-assisted generation constraints. These relationships are often treated as separate outputs, making it difficult to explain how perceptual evidence supports design generation. This study proposes an evidence-weighted heterogeneous graph mapping approach for analyzing symmetry–asymmetry patterns in electric motorcycle morphology design. A dataset of 176 electric motorcycles from 31 brands was constructed and organized into whole-vehicle archetypes and local styling modules. Kansei engineering and semantic differential evaluation obtained perceptual data from 78 valid questionnaires. The Analytic Hierarchy Process (AHP) was applied to determine expert-based criterion weights, and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) was used to rank whole-vehicle and local-module alternatives. In the proposed graph, vehicle samples, styling modules, Kansei dimensions, evaluation weights, ranking outputs, and prompt constraints are represented as heterogeneous nodes, while perceptual association, module–whole coordination, symmetry–asymmetry balance, and evidence-to-generation mapping are represented as weighted edges. The results identify system integration, form proportion and tension, safety perception, visual futurism, and brand identity as dominant perceptual dimensions. A second-round evaluation with 71 participants assessed the correspondence between the generated concepts and the target perceptual semantics. The proposed graph mapping provides an explicit and traceable evidence path from perceptual evaluation and multi-criteria decision support to AI-assisted concept-generation constraints.