自适应过滤算法在社区E-learning的个性化服务系统中的研究

Research on a personalized knowledge service system for community E-learning using an adaptive filtering algorithm

针对学习型社区中的教育需求,在传统算法上加以改进,提出了一种基于向量空间模型的教育资源自适应过滤算法.通过训练算法,提取特征向量和伪反馈建立初始模板,设置初始阈值.然后通过过滤算法根据用户的反馈信息自适应地调整模板和阈值.该算法在执行过程中,不需要大量的初始文本,同时在过滤的过程中可不断地进行自主学习来提高过滤精度.该算法已在个性化知识服务系统中进行验证,结果表明是有效的.

To effectively provide personalized E-learning in a community, an adaptive filtering algorithm for identifying appropriate teaching resources was developed. It is based on a vector space model, an improvement on traditional algorithms used for this purpose. Firstly, feature selection and pseudo feedback were used to establish the initial templates and thresholds through a training algorithm. Then the user＇s feedback was utilized to modify the templates and thresholds adaptively for the filtering algorithm. The algorithm did not need massive quantities of initial texts to begin the process of filtering. Furthermore, filtering precision improved during the process through self learning. The algorithm proved effective as a personalized knowledge service system for community E-learning.