Research On P2P Recommendation Method Of Online Educational Resources Based On Learner Social Relationship Analysis

With the introduction and popularization of the new model of online education,more learners have the opportunity to visit and learn on the online platform.Correspondingly,the platform has accumulated massive teaching behavior data and knowledge resources to provide updates and improvements to the platform itself However,the existing online education model and education platform still have many shortcomings.For example,the existing online education models and education platforms are centered on education platforms or educational resources,resulting in limited educational resources for learners to choose from.Moreover,learners are faced with inefficient retrieval of educational resources,and difficulties in matching educational resources to the needs of high-quality educational resources,causing the resources not be fully utilized.Additionally,the existing online education platform cannot effectively integrate offline education resources and services to provide a comprehensive overall learning plan for learners.Finally,the existing online education platform cannot provide accurate personalized services,and interactivity is lacking for learning in traditional online education platforms.In response to the above problems,this paper proposes a new type of smart education model,that is,the learner-centered smart education model.Based on this new model,in the application scenario,we consider the similarity of learners and the analysis of the social relationship between them,to recommend learning resources.While optimizing the traditional platform for recommending learners,it emphasizes the sharing and recommendation of learning resources between learners,that is,the P2 P recommendation mentioned in this study.Specifically,this paper provides a feasible method for automatically constructing learning groups.By determining the similarity between learners,the learning groups are divided.In addition,by using the representation method of the multi-dimensional social relationship graph model,the complex interaction relationship is expressed and described,on this basis,authoritative learner discovery and learning partner recommendation are performed.Finally,the research integrates clustering of similar learners and analysis of social relationships,adopts the factorization machine model to recommend learning resources for learners,and gradually improves its generalization ability and optimizes the recommendation results through model training.For each part of the research content,this article uses the real data set provided by the research group.After preprocessing,the experimental effect is verified,and the corresponding contrast experiments are designed to visually display and analyze the performance of the model,and also consider the influence of the active learners on the application effect of the model.The final verification shows that the learning group division and social relationship analysis methods proposed in this paper are fully reasonable.The recommendation model proposed in the study has obvious improvements compared to the traditional models,thus demonstrating how to support the realization of the learner-centered smart education model.Finally,the research method is applied,that is,the design and implementation of a preliminary new online education system are displayed.