【作者】 尤天舒；

【导师】 钟绍春；

【作者基本信息】 东北师范大学 ， 软件与理论， 2007， 硕士

【摘要】 随着计算机技术与通信技术的发展,我们己经步入了一个信息经济时代,信息技术在各个领域的应用正飞速的改变着人们工作、生活和学习的方式。远程教育的发展面临着前所未有的挑战。在远程教学环境下,教与学时空分离、学生个别化自主学习等特点使得教学管理者对学生进行评估时缺乏有力的数据和依据,其原因在于大多数现有的基于Web的远程教育系统缺少对学习过程的监控以及对学生在线学习行为的评估。当前的许多网络教学平台不同程度的存在着一些问题,主要表现在教学方法单一,教学双方互动性差等问题,因此现有的个性化学习平台并不能很好的解决个别化学习的需求,所以也就无法对学习者实施很好的个性化的学习服务,学习者无法针对自己的不足及时进行调整。针对这一问题本文提出了基于数据挖掘技术的个性化学习平台,个性化学习平台主要针对物理学科的学习进行评估。由于数据挖掘技术能够从海量的数据中发现一些未知的、有价值的规律,无疑为个性化的教育服务提供了强有力的支持。本文对决策树分类算法进行了分析和研究,认真分析和对比了ID3算法和C4.5算法的优缺点,引用了一种基于属性相关性的C4.5决策树规则简化算法。并将其应用到了物理学科个性化学习评估中,并通过实验数据证明评价方法的优劣。更多还原

【Abstract】 According to the development of computer and communication technique ,it is a age of information and economy, people is changing their ways of lives by the application of information technology in each field .The development of distance education is facing a challenge. The evaluation of the students’learning is very important through teaching process. Under the distance education environment, teaching is separate to learning and the students learning by themselves, all these characters make the teaching supervisor feel lack of useful data when evaluating. This is because of the lack of the monitoring on students’learning process and the evaluation of the students’behavior online by the distance educational system based on web. Nowadays, there are many kinds of problems in many educational environments based on web, primarily are the lack of methods on teaching and communication between teachers and students. What we have gotten for personalized learning environment can not solve the needs on personalized learning, thus these can not serve the learner very well on personalized learning, also the learner can not adjust themselves in time according to their shortcoming. To solve this problem, we present a environment model of personalized learning based on data mining technology, especially for the evaluation online of physics.Since we can find many unknown and useful regulations through using data mining technology, it is undoubtedly giving the personalized education service powerful support. Here we analyzed decision tree classification algorism, compared the ID3 algorism and the C4.5 algorism. Though the study on the C4.5 algorism, we used a easier algorism of C4.5 decision tree regulation based on attribute correlation which is already existed, and applied it into the individualized learning evaluation .We testified it by experiment data.更多还原