Research And Implementation Of The Recommendation System Based On Educational Resources

With the deepening application of information technology in educational field,digital education has become an important part of modern education.The educational resource platforms satisfy the growing demand for educational resources,meanwhile,changes the way people acquire educational resources.However,with the arrival of the era of big data,the rapid growth of the number of educational resources results in the problem of "information overload".People have to spend a lot of time and energy to find and choose the most suitable educational resources,which seriously reduces the efficiency and resource utilization.Therefore,it is of great significance to solve the problem that it is difficult for users to find what they want from massive resources when they use educational resources platform.Based on the above background,this thesis analyzes the characteristics of educational resources,researches on general architecture of recommendation system and the relevant theory and technology.On the basis of the current research results,this thesis designs and implements a recommendation system based on educational resources.The system uses four-tier layers of architecture,from top to bottom in order to user UI layer,recommendation layer,offline layer and data layer.The user UI layer is responsible for the interaction between the user and the system.The recommendation layer implements the personalized recommendation.The main function of the offline layer is to calculate and analyze the data.The storage layer provides data preprocessing and storage.Among them,this thesis focuses on the design and implementation of the recommended algorithm and function on recommendation layer.The main contributions of the thesis include:1.In the aspect of algorithm,the principle of association rule algorithm and cooperative filtering algorithm are deeply studied.Then this thesis analyzes their shortcomings,and makes some improvement.Based on the Eclat algorithm,this thesis optimizes the efficiency of its frequent itemsets.In addition,a collaborative filtering algorithm based on user and item combination is proposed,in which the factors such as sparseness,item characteristic attribute and user attribute preference are added to improve the accuracy of forecasting score.2.In the aspect of function,two above-mentioned algorithms are applied to each recommendation function.This thesis designs and implements the following functions,namely,download the recommendation,rating recommendation,personalized search and mailbox push function.3.The two improved algorithms are experimented and compared with the traditional algorithms.The results show that the proposed algorithm is better.In addition,the function modules and the performance of the system are tested,and the test results have reached the desired goal.In this thesis,a recommendation system based on educational resources is successfully implemented,and the system has good practicability.