Laboratory-Based Dataset for Cognitive Load Prediction Across Different Map Types​

30 participants（15 females and 15 males）were recruited. All participants were master’s students. They all had normal naked eye vision or corrected vision, and were right-handed. They had prior education in map reading and frequently used maps, but they had no prior knowledge of the map content involved in the experiment.Participants were assigned to three groups based on the Santa Barbara Sense of Direction Scale self-assessment, ensuring that each group had a roughly equivalent number of participants, gender ratio, proportion of geographic education, frequency of map usage, and sense of direction.The experiment was conducted in a laboratory setting. The experimental display was a Dell desktop computer with a 23-inch screen and a resolution of 1920×1080 pixels. The eye-tracking data was collected using aSeePro desktop eye-tracking device with a sampling frequency of 140 Hz. The aSeeStudio software was used to manage and analyze the eye-tracking data.We selected a 3D map from Amap, along with 2D and imagery maps depicting approximately the same geographical area , and ensured identical spatial coverage across all map types through a preliminary questionnaire survey. These three map types served as distinct experimental materials. Participants, pre-grouped based on the Santa Barbara Sense of Direction Scale, were assigned to one of the three map conditions. During the experiment, participants were given six seconds to freely explore the map. Subsequently, they performed tasks following audio instructions. The target location name was displayed on the map alongside a red bounding box, and participants were required to select the specified geographical area by clicking on it.