Research And Implementation Of Personalized Learning System Based On Cloud Platform

With the development of Internet and cloud computing,the way people learn has been changing in the mode of "Internet plus education",which leads to a concept of wisdom education being put forward and affects the way people learn.In recent years,there are more and more network courses,so a large number of learners begin to study on the network,which is convenient for learners to acquire knowledge.However,there are also some problems.For example,there are abundant resources on the website,which leads to the phenomenon of the trek phenomenon of the learning process,repeated construction of resources and a low rate of resources reuse.It is crucial to solve these problems effectively.As the data of learner behavior are stored in the process of learning,the processing reasonably of the generated data can effectively promote the development of educational information.Because of the difference between different learners in learning style,cognitive level and learning preferences,there is a different understanding for the same teaching resources.How to optimize teaching resources by using these data effectively,to improve the efficiency of learning,and ultimately to achieve the purpose of personalized teaching have become an important research field.By introducing the research status of cloud platform,educational cloud and personalized learning,this paper analyzes the advantages of HDFS and MapReduce in the Hadoop environment of cloud computing mainstream platform,such as storage,scheduling,and data analysis of educational resources.The development of Hadoop technology provides a new train for the realization of personalized learning,which applies learning analysis technology to deal with a large amount of data in teaching process.For a large number of teaching resources and diverse learners,this paper builds a platform of Hadoop cloud to develop the personalized learning system,collects and stores a large number of learners' behavior logs by using the advantages of distributed technology and cloud computing.And,the personalized recommendations are made for different learner,which optimizes the process of teaching and learning and enhances the learners' interest and efficiency.Due to students differ in their learning styles,cognitive abilities and attention resources,this paper initially investigates the learner's preferences through the Felder-Silverman learning style scale and the pre-test of the cognitive level,and builds a more accurate learning model by learning analysis techniques to deal with behavioral imprinting.Taking into account the differences among learners and the possible changes at any time but the difference among teaching resources is relatively stable,this paper chooses collaborative filtering recommendation algorithm based on projects,takes into account the similarity of resource attributes,and transforms user's behavior into scoring to recommend personalized learning content for learners.Because the relationship between learning resources and learners are interrelated,it provides the optimal learning path for students through the analysis of the relationship of ant colony algorithm.Based on the learning system in cloud platform,it collects earning behavior logs by using the Flume tool,and creates a personalized learning space for learners through learning analysis technique.