Dynamic perception graphs from spatial behavior data in Minecraft-based city environments

Dynamic Perception Graphs from Spatial Behavior Data in Minecraft-Based City Environments This dataset contains dynamic perception graphs derived from spatial behavior data of 136 participants exploring 10 city locations modeled after Daejeon, South Korea, within a Minecraft-based virtual testbed (see [1] for environment details). Participants explored the environment for approximately one hour, during which their position and orientation were recorded every second. Perception graphs were generated following the methodology described in [2], using a visual angle of 120° and a visual range of 10 meters. These graphs were segmented into 30-second intervals to create dynamic perception graphs as described in [3]. In total, 12,911 graphs were generated. This dataset is primarily associated with [3], which presents the detailed methodology and findings. Researchers using this dataset are encouraged to cite [3]. References: [1] Chang, Mi, et al. "The influence of virtual tour on urban visitor using a network approach." Advanced Engineering Informatics 56 (2023): 102025. [2] Lee, Gi-bbeum, et al. "Perception graph for representing visuospatial behavior in virtual environments: A case study for Daejeon City." Advanced Engineering Informatics 62 (2024): 102594. [3] Lee, Gi-bbeum, & Lee, Ji-Hyun. "Graph approach for Gibson’s ecological optics with dynamics of network motifs". Advanced Engineering Informatics 69, (2026): 103865.