Research And Implementation Of User Portrait System Based On Weibo Data

With the gradual arrival of the 5G era,the Internet industry is brewing a new round of outbreaks.At the same time,the Weibo social platform that carries people's wishes has also developed rapidly.Users post a wealth of personal information and massive personal news on the Weibo platform,and these data are all self-published behaviors of the user,with a high degree of credibility and diversity.By collecting,integrating,and analyzing these diverse data generated by users,it is possible to depict user portraits as comprehensively and accurately as possible.This thesis mainly collects user data from the Weibo platform,analyzes and mines user characteristics,and constructs a relatively complete user portrait.Aiming at the problem of the separation of the label model and the portrait generation in the common user portrait construction process,this thesis uses the fusion method of the label generation and the portrait generation,and proposes a user portrait generation model based on WBC fusion,which further improves the user portrait compared to the separation method accuracy.The main work of this thesis is as follows.Firstly,collecting all relevant data of users on Weibo platform based on crawler technology.Using technologies such as automatic login,account pool and proxy IP pool,to achieve a daily million-level data collection module.Using data pre-processing technologies such as clustering algorithms and data cleaning methods,to build and maintain a regular,authentic user information data set.Secondly,based on the topic mining model,the completed preprocessed text is modeled to obtain all potential user interests.And the microblog user-defined interest tags,collected by the data collection module,are analyzed mathematically and statistically.A method of combining the two parts is proposed to build a more comprehensive labeling model.Thirdly,using user data sets and label models,a user portrait generation model based on WBC fusion is proposed to achieve the goal of retaining both word level and text level features,and merge the label model and portrait generation in the process of user portrait construction,and improve the accuracy of user portraits.Finally,design and implement the user portrait system,including user management and visual operation of specific functions,and design methods to improve response time,reliability and security of the system.