Research On Learning Genre Combining With Exercise And Its Content Recommendation

In recent years,the development of education in China has stepped into a new stage of attaching importance to students physical and mental health,and promoting students personalized development.DGBL(Digital Game-based Learning)is an emerging form of education.It combines serious learning process with digital interactive entertainment and adopts the idea of entertaining,which makes the learning process no longer boring,and encourages children to be more interested and motivated to learning curriculum.The main interaction methods includes PC-side interaction,mobile device-side interaction,and VR device-side interactionHowever,there are three problems with the existing DGBL:First,there are hidden health hazards in the interactive mode.Interactions on a PC or mobile device have the problems of sedentary and long-term staring at the electronic screen to damage vision;interactions on VR devices often have VR sickness,and VR manufacturers do not recommend children under 13 to use.Second,there is insufficient support for the long-term factor in learning activities.In DGBL-related work,it has been mentioned many times that the longitudinal study of learners should be analyzed,indicating that the consideration of long-term factor in learning behavior is necessary.However,part of the DGBL application does not record the identity of the user,and some applications record but just simply recording the progress of the course,without further collecting learning data and analyzing them.Third,the individual differences and long-term development of students are ignored,and personalized and adaptive learning content is not well provided.Some DGBL applications have not yet based on long-term historical learning data to provide and adjust appropriate learning content for each student according to their ability level and ability changes.In order to promote healthy,long-term,and personalized learning for students,This paper presents a new learning genre combining with exercise-Exer-Learning,and presents its content recommendation method.Exer-Learning combines the three aspects of learning,exercise and game fun,and provides a healthy and harmless way of learning for learners.Exer-Learning takes the long-term factors in DGBL learning activities into consideration,supports long-term learning and exercise,and records data during this process.And it uses student historical learning data to provide everyone with adaptive learning content,and provide a personalized learning experience.Specifically,the main work and contributions of this article are summarized as follows:1.This research presents a learning genre—Exer-Learning,which combines learning activities with physical exercises in the context of fun games.To exemplify such genre,we design an interactive instance,which integrates physical exercise as a key element in a language learning game.By naturally using location and body movement in the projection area as the way of interaction,learners can learn knowledge and carry out certain physical activities at the same time.We conducted a controlled study to understand the optimal design tradeoffs of two elements(level of physical exercise and knowledge provided by the game)under this Exer-Learning genre.Four trade-offs conditions are distinguished and their effects on participants'exercise benefit,learning efficacy,and fun of the game were examined.Results show that the benefits of exercise and learning are relatively high under the condition of high physical exercise and low knowledge.However,high physical exercise conditions reduce participants' fun in the learning.2.This research implemented an adaptive learning problem pushing strategy for English learning content,which can be divided into two parts:learner modeling and model application.First,we use the Deep Knowledge Tracing(DKT)model implemented by Recurrent Neural Network(RNN)to model our learner's knowledge state,on student simulation training data.Then apply the trained learner model in adaptive learning tasks to present each one with adaptive learning content,and provide a personalized learning experience.Finally,we evaluated the overall adaptive learning according the system criterion,academic criterion,and satisfaction criterion.The experimental results show that the accuracy of the student model is relatively high,the adaptive learning group is better than the control group in terms of learning performance and learning efficiency,and the score of subjective satisfaction is also desirable.