Research On Data Mining And Visualization Based On WeChat Public Platform

In the Internet age,a lot of information emerges every day.It is impossible for the users to choose information they interested by themselves.Therefore,a recommended system which can automatically and intelligently filters out useless information and then recommends some interesting or useful information to Internet users is very necessary and important.This kind of ideal recommended system is thus studied in this paper.In addition,a prediction recommendation algorithm toward to the latest published article is proposed and used in the recommended system to filter and recommend interesting or useful information for the users.The specific research work of this paper is as follows:First,we build up a Web framework of the recommendation system by using HTML,and improve the web page through CSS.The JavaScript is used here to increase the dynamic function of the page.To achieve as well as improve the beauty and practicality of the page,D3.js is used in the data icon part of the page.Secondly,we write a web crawler by using python to crawl the background data of Wechat public platform.The web crawler is a kind of script that can automatically crawls information from the web.The data,such as the reading number of each article,the liking number,forwarding and comments and so on,are obtained by utilizing the crawler.Both of these information and the interaction data are stored in a MySQL database for further use.Thirdly,a prediction recommendation algorithm toward the latest published article of WeChat public platform is presented.By analyzing the potential of the article and the characteristics of other articles wrote by the same author,predicted rating and intelligent sorting can be obtained by using the proposed algorithm.Based on these results,we can then recommend the most explosive potential articles and the readings which have already been read by more than 100 thousand times to the public to ac hieve personalized recommendations.Simulation results demonstrate that intelligent recommendation can be achieved through using this new recommendation algorithm.Finally,an intelligent recommendation system which contains thousands of WeChat public number is built up.The system can help users to search and get different information according to their requirements.It can also be used to predict and recommend the explosion point article,which will greatly saves the user's information retrieval time.