The Research Of Multi-dimensional Learning Community Based On Clustering Algorithm In E-learning Environment

With hit of the third wave of the Internet, people are increasingly dependent on the network. E-learning, as a life-long learning way, together with the virtual communities, is changing learning and life styles of the hundreds of million Internet users. The virtual learning community is an important application mode of learning environment, which combines the advantages of E-learning and virtual communities so as to attract the attentions of a growing number of educational institutions and Internet companies.This paper proposes a new type of multi-dimensional learning community. As opposed to the traditional Virtual Learning Community, by assessing and classifying students through their learning status real-time and dynamic, the new-type community automatically divides students into different levels of groups for group learning. In the design and implementation of assessing and classifying students, we collect their learning progress data out of a multi-mode interactive Ontology-based learning environment and generate a cognitive diagnosis model which can get a correct feedback of their learning states. After comprehensive analysis, we select a reasonable and effective clustering algorithm to do cluster analysis to the model. On this basis, we successfully implement a new data mining application in E-learning environment.The research work of this paper includes the following components.Firstly, this paper conducts a research on current situations of mainstream learning communities on the Internet, analyzes their characteristics and summarizes the existing deficiencies. And then through a set of experiment analysis to a group of target users of the virtual learning community, we discover the differences in their learning levels, needs of knowledge, experience and interests, so that we prove the necessity of building learning community group of different levels. Based on the conclusion above, we propose a new concept of virtual learning communities and describe the design ideas of its multi-dimensional learning environment.In the second part, relying on the current course center platform, we verify the belief that it is feasible to produce a cognitive diagnosis model of learning status by means of collecting learning progress data in Ontology-based learning environment. In the summary of our model-generated results and the analysis of the simulation results from other researchers, we systematically and carefully compare the cluster algorithms related, and then select the ideal algorithm, so that we are able to carry out a more reasonable student classification.In the third part, we have implemented the proposed new-type learning community and made its deployment. We realize the function that through cluster analysis, registered students can be automatically divided into groups with different levels. In addition, we integrate the learning community into the existing education platform of the School of continuing education of Shanghai Jiaotong University, to prepare for its future utilization in actual teaching and further research.Finally, we summarize the work done in this paper, and look forward to the next stage of our work.