Learning Immersion Assessment Model Based on Multi-dimensional Physiological Characteristics

• Published in: 2020 IEEE International Conference on Power, Intelligent Computing and Systems (ICPICS) pp. 87 - 90
• Main Authors: Wan, Boxin, Guo, Junqi
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2020
• Subjects: Biomedical monitoring; Brain modeling; Electroencephalography; Feature extraction; learning immersion; physiological characteristics; Physiology; Support vector machines; SVM; Training
• DOI: 10.1109/ICPICS50287.2020.9202208

The wisdom classroom full of immersion experience could create an excellent sense of learning experience for learners to improve their learning effect, and the level of learning immersion has become one of the indicators to assess students' learning state. However, the traditional way to evaluate learning immersion is greatly influenced by the subjective factors of individuals. This study proposes a novel method to assess learning immersion based on physiological characteristics: by constructing two learning scenes with different immersion senses: VR video watching (for high immersion) and normal document reading (for low immersion) to induce immersion. During the learning process, the subjects' PPG signals and EEG signals were collected for later preprocessing and extracting. By entering different compositions of feature vectors to train SVM classifier model and comparing their prediction accuracy and training time, the results show that the most suitable feature vectors to assess learning immersion are: pulse rate, the ratio of attention score and relaxation score, high alpha wave of EEG. The precision of the model reaches 88.93%.