Research On Important User Recommendation Methods Based On Personalized Tag And Microblog Topic

With the rapid development and popularization of the Internet and smart phones,social networking platforms have become hot.And more and more users are relying on social networking platforms to communicate in real time and obtain relevant hotspot information.The most favorite user in social networking platforms is microblog.Because users can publish information as well as deliver information in microblog.But at the same time,this will lead to a rapid increase in the amount of information in the network and bring about information overload.However,these overloaded messages contain a lot of useful information.In order to extract useful information from these overloaded information,personalized recommendation algorithms and technologies have emerged.For the microblog platform,recommending important users in relevant areas of interest for users can effectively increase the user's reliance on the platform.In the microblog platform,since the user has just registered to a circle of friends with a certain size,the different stage of the user makes the effective information displayed differently.This theme analyzed the user's effective information at different stages and conducted research on important user recommendation methods based on personalized tag and microblog topic.It mainly consists of two parts.On the one hand,In order to solve the problem of personalized label sparseness,this theme increases the number of labels by segmenting the personalized labels.In order to solve the recommendation accuracy low problem,this theme also gives a label weight setting scheme and combines the relationship between users and the importance of users.Through the solution of the above two problems,an important microblog user recommendation algorithm based on personalized labels is proposed.The algorithm effectively solves the sparse problem and improves the recommendation accuracy.On the other hand,In order to solve the problem of low recommendation accuracy,this theme proposes an important microblog user recommendation algorithm based onsimilar topics.The algorithm firstly improves the HITS algorithm,then uses the improved HITS algorithm to classify the user category and accurately calculates the user authority and centrality.Finally,the calculation of topic similarity,authority and centrality are introduced to improve the topic similarity between users,so as to effectively improve the recommendation accuracy.In order to prove the validity and accuracy of the method proposed in this theme,comparative experiments were conducted.The experiments are divided into two parts.The first part was a comparative experiment on the microblogging dataset acquired by crawler technology.Through the experimental results,we could see that the important microblog user recommendation algorithm based on tags proposed in this theme can effectively solve the problem of personalized label sparseness.At the same time,it combined the similarity of the label with the importance of the user to improve the accuracy of the recommendation.The second part was also acquiring the dataset through crawler technology and performing comparative experiments on the dataset.The experimental results show that the proposed microblog user recommendation algorithm based on similar topics can improve the accuracy of user authority and centrality calculation and the accuracy of topic similarity calculation between users,and effectively improve the recommendation accuracy.