改进的E-learning社区自组织算法  

Improved Self-organizing Algorithm in Constructing E-learning Communities

作　　者：金晶[1] 陈清华[1] 罗恒[1] 

机构地区：[1]上海交通大学计算机科学与工程系,上海200240

出　　处：《计算机工程》2008年第19期266-268,共3页Computer Engineering

基　　金：国家自然科学基金资助项目(60372078)

摘　　要：为了解决E-learning过程中缺乏对学习者有针对性指导的难题,该文提出一种构建E-learning社区的改进算法。该算法采用P2P的架构,通过改进的Hebbian学习法则不断加强具有类似兴趣的学习者之间的联系,从而自动调整社区结构,最终将具有类似兴趣的学习者组织在一起。实验结果证明,与传统的P2P架构的Hebbian学习算法相比,该算法达到了更好的社区建设质量和更高的社区建设速度。To solve the problem that learners are unable to get personalized instruction, an improved algorithm is proposed for constructing self-organized communities. This algorithm adopts P2P architecture, and strengthens connections between learners with similar interests based on improved Hebbian learning law. In this way, community structure is adjusted automatically, organizing learners with similar interests together eventually. Experimental results show that compared with traditional Hebbian-based P2P algorithm, this algorithm achieves better construction quality and higher construction speed.

关 键 词：Hebbian学习 P2P架构 E—learning社区 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]