Design And Implementation Of Collaborative Learning System Based On WGDK-means

With the continuous development of the "Internet +",online learning has become more and more popular.However,currently,most of the online learning platforms only provide learners with learning data and simple message board functions,which is impossible to provide a grouping cooperative learning environment for learners.It leads to the phenomena as follows: the learners are in lack of effective communication;the learning efficiency is very low;the course completion rate is very low.Therefore,establishing group cooperative learning environment is an effective means to improve learning efficiency.In this way,how to divide groups involves the clustering algorithm.In order to improve the group-division quality,this thesis proposes a weight-based meshing density peak K-means(WGDK-means)algorithm,which combines and improves the two algorithms(the K-means algorithm and the density peak algorithm)according to their advantages and disadvantages.Finally,it applies the WGDK-means algorithm to the learner model.The main work of this thesis is as follows:(1)It proposed the WGDK-means algorithm.Firstly,it concretely analyzed the advantages and disadvantages of the K-means algorithm.In view of the disadvantages that the K value needs to be given in advance,and the initial center point is sensitive to the selection of the initial center point,it put forward the notion that the K initial points are obtained by density peak algorithm in advance,and then K-means clustering is carried out.However,due to the high time-complexity of the density peak algorithm,it is not favorable for the direct usage.Therefore,it proposed an improved mentality of obtaining K initial points on the basis of the meshing density peak algorithm.Then it elaborately introduced the algorithm idea as well as the implementation steps of WGDK-means.Finally,through the experiment,it verifies that the average accuracy of the WGDK-means algorithm is improved by nearly 8% in comparison with the K-means algorithm in the learner's characteristic data centralization.(2)It proposed the gradient grouping algorithm on the basis of the WGDK-means.Firstly,it applied the WGDK-means algorithm to the learner's data centralization,and established the learner's characteristic models,including the knowledge-level model,the cognitive-level model,the learning preference model,and the cooperative ability model.Finally,it applied the WGDK-means algorithm to the cooperative learning grouping,and put forward the gradient grouping algorithm on the basis of the WGDK-means,and proves by the experiment that the gradient grouping algorithm has a good practical application value in learners' cooperative grouping.(3)It completed the design and implementation of the cooperative learning system.According to the requirement of the cooperative learning system,this thesis applies the gradient grouping algorithm of WGDK-means to the cooperative learning system.Through the design for the system structure,the function,and the database,it applied the current popular development technology to establish the cooperative learning system and put it into the practical utilization.