Design And Implementation Of Personalized Recommender System For Online Learning Resources

The development of education informatization has brought about major changes in the form of education and learning methods.According to the 12 th Five-Year Plan of the Ministry of Education,the existing education network and campus network will be upgraded to education informatization,and the online teaching platform is the main way of showing smart teaching.Under the advocacy of national policies,various online learning platforms have also been launched one after another,which bring convenience to learners but also cause troubles.First,learners are faced with a large number of online learning resources with varying contents,and they can’t quickly find learning resources that meet their own learning needs and preferences.Second,most learners can’t fully understand their own learning situation,don’t know where to learn and how to learn.Based on the above background,this thesis aims at the overload phenomenon of learning resources in the field of online education,in order to solve the problem of learners’ "information trek",a recommendation algorithm is introduced into the online learning system to provide learners with personalized learning content.Firstly,we analyzed the relevant background and research status of online learning resource recommendation,and the existing problems are improved and optimized.Taking college students as the main research object,and the relevant knowledge points of high-frequency electronic circuits as the learning resource library,the design and implementation of A personalized recommendation system for online learning resources is developed to recommend personalized learning resources that are more suitable for learners’ needs and preferences,and improve learners’ learning efficiency and user experience.The main research work of this thesis is as follows:1.Build a user profile model.The learner is profiled through the learner’s personal basic information and learning behavior data,and the theoretical model of the labeling system of the user portrait is constructed.The main tasks are: data collection and label the relevant data;build a user profile model,calculate the learner’s behavioral weight on each learning label,and vectorize it.2.Analysis system requirements and design overall scheme of online learning resource recommendation system.Firstly,analyzed the requirements from function aspect and performance aspect.Secondly,designed overall scheme of the system through the requirements analysis,including the front-end Scheme and back-end scheme,designed the relevant functional modules of the system.Finally,designed database through the relationship of main entities,including the detailed structure of the data table,the main fields and the relationship between the tables.3.Design,implementation and improvement of learning resource recommendation algorithm.In this thesis we used the user-based collaborative filtering algorithm and improve its defects.Firstly,we choose Pearson similarity as final calculation method of student similarity which is integrated resource weight factor and score difference factor.And then designed a collaborative filtering algorithm integrating user portrait and K-means clustering algorithm,and compared improved recommendation algorithm with the traditional algorithm on the relevant data set,calculated the evaluation indicators such as RMSE,MAE,precision,recall and F1.4.Implementation and test of personalized recommendation system for online learning resources.This thesis designs and implements each functional module of the system in detail,and displays the system implementation interface.In the above,formulated a comprehensive test plan and wrote test cases,each module was tested separately.The results showed that the functionality of the system was in line with the expected results.