Personal Educational Resources Push Method Research And Application Of "Xiang Jiao Yun"

Personalized information push can effectively solve the information overload,and can help users quickly and easily mining the useful and interest-related information from the massive information,in recent years,it has been widely used in the field of e-commerce,researchs but also tend to apply personalized information recommendation to the field of education,however,between the field of e-commerce and education has a greater difference,in the education field,personalized information promotion is still in the exploration phase,can not directly applies the more mature electronics business technology to the field of education as simple,before reasonablely use it,should fully consider the characteristics of educational resources.To this end,this study takes full account of the particularity of the field of education,based on the domestic and international research status and results,combined with the field of e-commerce"personalized recommendation","data mining" technology to achieve personalized education resources push.This paper mainly includes the following contents:(1)Analysis and design of educational resources and user interesting model construct.By analyzing the relationship between user and educational resources in the promotion of personalized educational resources,this massage designed a method based on LOM to classify the educational resources and construct its model;for the analyze and design of constructing user interesting model is divided into three aspects:acquisition,representation and renewal.,use the method of explicit and implicit to build the model of user interesting.(2)Design of personalized push system.The whole system hierarchical design,divided into the metadata management layer,personalized push service layer,educational user interface layer,metadata management layer is responsible for the construction of educational resources,personalized push service layer is responsible for the operation of resources push,education user interface layer is responsible for the collection of user information behavior data and user interaction.(3)Push algorithm design and testing.Desiged three kinds of push algorithm:User-based algorithm based on Spark and HDFS for the large data background,this algorithm can deal with the large datasets qucikly and efficiently;the optimization of Item-based algorithm for educational resource characteristics,this algorithm could obviously improve the accuracy of recommendation;the real-time recommendation algorithm and model based on Spark mllib and Streaming,this model can accept behavior data of user and recommend corresponding resources in real time.Meanwhile,validated and tested experimentally for the corresponding algorithms.In this paper,The method of user interesting model and description&classification of educational resources can achieve a good correlation between educational resources and user relations,based this designed the overall framework and implementation process of personalized information push system of educational resources,it is a good solution for educational recommendation.In addition,the recommended algorithm for the analysis of large data and educational characteristics has been verified can improve the quality of recommendation.