Comprehensive design thinking is vital for architects to resolve spatial conflicts in architectural design. To address the limitations of traditional brainstorming (TB) in complex constraint scenarios, this study proposes a cognitive science-based semantic network simulation (SN) method. A controlled experiment was conducted with 60 architecture students randomly divided into SN and TB groups, tasked with solving the same rural cultural center entrance design conflict within 10 min, with eye-tracking monitoring subconscious thinking. Multi-dimensional evaluation shows TB generates more proposals, but SN outperforms TB in innovation, conflict resolution, and thought explicitness. For architects, SN provides a structured thinking tool: decompose design conflicts into nodes, establish logical connections, and generate innovative solutions through cross-domain association. The findings offer actionable methods for architectural design practice and education, helping break the “function–innovation–efficiency” trade-off.
Dong et al. (Mon,) studied this question.