基于Web使用挖掘的个性化学习推荐系统

针对当前E learning推荐系统存在的问题，引入多Agent（MAS）系统，提出一种基于Web使用挖掘的集成MAS与Web services的分布式智能推荐系统模型。该模型能动态生成基于用户使用信息的个性化链接页面，有效地帮助学员找到所需的资源信息。提出了一种基于最近最少使用策略的系统推荐算法，包括系统整体实现算法、系统聚类算法及推荐算法。实时性能分析显示该系统运行性能良好。

Recommendation system based on Web usage mining for personalized E-learning

To solve the problems of current E-learning recommendation systems by adopting multi-agent system(MAS),a new distributed intelligent recommendation system based on Web usage mining(WUM) was put forward integrating MAS with Web services.It can help learners find resource requested by dynamically generating personal link pages based on the user identification.By adopting least recently used(LRU) policy,the recommendation algorithm of the system was proposed including the global implementing algorithm,clustering algorithm,and suggestion algorithm.Real-time application and experiments show that it is feasible and effective.