Design And Implementation Of Intelligent Adaptive Learning System Based On Multi-Feature Fusion Knowledge Tracking

With the advancement of Internet technology,online education has developed rapidly.In particular,due to the impact of the new crown epidemic in the past two years,the duration of online learning has increased significantly,and an efficient and intelligent online learning system has become an important demand.The intelligent adaptive learning system uses artificial intelligence algorithms to create an intelligent and diversified learning environment for users,which is a research hotspot in the field of online education.At present,the intelligent adaptive learning system still has problems such as low accuracy of knowledge state prediction and insignificant personalized learning effect,which affects the actual learning effect.Based on deep knowledge tracking and reinforcement learning theory,this paper improves the intelligent adaptive learning model,and designs and develops an efficient and stable intelligent adaptive learning system.The main research results and innovations achieved include:1.Based on the principle of autoencoder and the theory of deep knowledge tracking,a deep knowledge tracking model(DKT-MF)that integrates multiple features is constructed.The test results on the public data set show that the AUC is up to 0.84413,the Accuracy is up to 0.78828,and the Compared with the original DKT model,the improvement is 2.9%and 6.24%,respectively.2.Combining reinforcement learning and intelligent adaptive learning theory,a DDQN-based multi-objective personalized exercise recommen-dation model(DDQN-MPRM)is constructed.The test results of using DKT-MF to simulate the answer of the learner show that:compared with the random recommendation method,the relevance of the recomm-ended exercises of the DDQN-MPRM algorithm increases from 0.00756 to 0.02892,the difficulty fluctuation decreases from 0.20227 to 0.03138,and the reward value is more stable and maximum 20%increase.3.Completed the detailed design,development and testing of the intelligent adaptive learning system.The system implements the functions of knowledge status evaluation and personalized exercise recommendation,with an evaluation accuracy rate of 89%,a recommended exercise knowledge point coverage rate of over 66%,and a knowledge point mastery rate of over 86%.