Research And Design On English E-Learning System For Schoolchildren Based On CDA And CDC CF Algorithm

With the popularity of Internet and intelligent terminals,children's English E-learning have undergone great changes.The learning resource has reached an unprecedented abundance.Most of the platforms,such as GuaGua,have adopted traditional CF(Collaborative Filtering)algorithms to implement the R&S(Recommendation and Search).However,the learning capacity of children is increasing rapidly,while the traditional CF algorithm cannot carry out real-time personalized recommendation based on the capacities.And the difficulty and applicability of learning resources are not well tagged and evaluated;the resources are unable to be confirmed whether suitable for the current learners.Combining with my research and experiments,this thesis proposed an improved hybrid CF algorithm based on the theories such as the CDA(Cognitive Diagnosis Assessment),the CDC(Content Difficult Coefficient),EKG(English Knowledge Graph)and FCT(Forgetting Curve Theory)to optimize the R&S algorithms of the platform.And this thesis also given the evaluation method on recommending effect.The main research is as followed:1,EKG and the model of the cognitive ability of the user were constructed;2,The difficulty evaluation model of resources was given,and the initial value was given;3,Constructed the user's learning process log database;4,Developed the dynamic matching algorithm based on user cognitive ability and learning resource difficulty coefficient.5,Developed a search and recommendation algorithm based on cognitive ability,CDC,EKG and FCT.According to the research of this thesis,the simulation experiments were carried out.The experiment shows that the improved CF can effectively locate the most matched learning resources.Moreover,the corresponding recommended result set is obviously reduced,and the accuracy,pertinence and applicability of the R&S in English E-learning platform are improved obviously.