Research On Course Recommendation Strategy Based On User Value Segmentation

In the context of open and shared knowledge,online learning platforms are developing rapidly,and people’s learning behavior is also changing.A large number of users login the platform and accumulate a large amount of data.It has always been a hot research area that how to analyze user value based on big data mining and how to provide personalized services for users.However,previous researches were mostly focused on traditional industries such as e-commerce and banking,while few researches were focused on online education.Based on the user data of an online vocational education platform,this paper conducts relevant user value segmentation and course recommendation strategy research.The main research is as follow:(1)Using improved RFM(Recency,Frequency,Monetary)model and the Entropy method to build a new user value segmentation model.First of all,combined with the user characteristics of the education platform,three user value indexes of course purchase,payment rate and learning duration are added on the basis of the original RFM model to form an improved RFM model.Secondly,by comparing DBI,Silhouette Coefficient and C-H values of the three clustering algorithms when k = 4,K-Means ++ algorithm is selected to separate user groups.Finally,the Entropy method is used to calculate the index weight.The weighted sum is used to get the value score,rank each user group,and form corresponding operation suggestions of the user group.(2)The course recommendation strategy model based on the Association Rules algorithm is presented.Firstly,the course data of all users and each user group are analyzed by the Association Rules algorithm respectively.Secondly,the results of each strong association rules are found and visualized.Finally,according to the results of the model,the optimal course combination is given,which is convenient for the platform to recommend the course combination package to users and improve user enthusiasm to purchase courses.(3)The course recommendation strategy model of the Item Collaborative Filtering algorithm based on improved cosine similarity is presented.First of all,based on the high quality user group obtained above,the course recommendation model is constructed by adopting the article collaborative filtering algorithm with improved cosine similarity.Secondly,by adjusting the parameter values in the model and observing the index results of the recommendation algorithm,the optimal parameters are selected for the platform users to recommend courses.Finally,this model is compared with other models based on the Item Collaborative Filtering algorithm,and the model in this paper has better indicators and better recommendation effect.The research of this paper is a method study of course recommendation strategy based on user value segmentation,and provides a reference case for online learning platform of vocational education.The corresponding personalized service and operation strategy mentioned in this paper have certain reference value.