Research On Intelligent Companionship Learning Experience For Primary Education Scenarios

Nowadays,parents attach importance to their children’s primary education but often lack time and correct pedagogical principles to accompany their children’s learning.Although artificial intelligence(AI)and augmented reality(AR)technologies have brought new possibilities for education,existing learning systems are still unable to perceive children’s emotional and learning state changes,which means that parents are still indispensable.They may also cause children’s self-control and cognitive problems due to smart devices such as mobile phones and tablets.In this work,we study how to integrate AI and AR into educational applications based on the classic ARCS theory to stimulate and replace the role of parents and provide companionship and heuristic education for children.We propose an intelligent companion learning system named Intelligent Augmented Reality Educator(IARE)for children to learn English words.The IARE realizes the perception and feedback of children’s engagement through the intelligent agent(IA)module,and presents the humanized interaction based on projective Augmented Reality.We allow children to interact with physical letters,thus avoiding the excessive interference of electronic devices.The main research contents are as follows:We propose an intelligent companion learning system named IARE,which seamlessly integrates AR and AI based on the ARCS model.It can accompany children to learn English words,stimulate positive psychological feedback in children through appropriate guidance,and improve children’s self-learning performance.We develop an IA module to imitate and substitute the role of parents in word learning scenarios of primary education.We propose an online lightweight learning engagement evaluation model that can achieve faster online evaluation based on the temporal multiple instance attention module proposed by Ma et al.and train it through self-collected and labeled video datasets.We propose a stability algorithm to schedule the character recognition network to improve the efficiency of character recognition,and improve the robustness of the model to diverse character recognition backgrounds and character shapes by synthesizing a dataset.IA can perceive the changes of children’s learning engagement and spelling status in real-time to adjust the learning process attributed to the algorithms and datasets mentioned above.We implement a learning scenario called exploring mode in our system,which induces children to conduct discovery-style unintentional learning and stimulates their interest and motivation.Combined with the other three modes and IA adjustor,children can maintain their learning engagement.We conducted a pilot study to test the efficacy of our system with the traditional flashcards group as the control group and 14 third-graders as participants.We collected qualitative and quantitative data to analyze the research results.The results show that our system can significantly improve children’s intrinsic motivation and self-efficacy.