Research And Application Of Topic Discovery Method Of Online Public Opinion On Food Safety

Food safety is related to people’s health and life safety,relevant public opinion has always been the focus of public concern.In the rapid development of Internet,public opinion on the Internet about food safety,easy to spread and expand quickly.With the increase of participants,public opinion continues to ferment into large public opinion events,which will bring great impact on social stability.Therefore,the research of online public opinion on food safety has important research significance and application value.In this paper,topic discovery of online public opinion on food safety is selected as the starting point,and the crawling technology,topic discovery model and system development technology of Online public opinion on food safety in Chinese video media are studied in depth.The main work is as follows:(1)Research on data crawl technology of food safety network public opinion.In the face of the difficulty of data acquisition and unstable crawling of dynamic web pages,this paper analyzes and studies the Ajax technology used by dynamic web pages,and adopts Selenium automation and reverse analysis to crawl bilibili video website about food safety video comments by combining Scrapy crawler framework.The operation performance of crawl technology and coping strategy of crawl limitation are compared and analyzed,which provides a reliable method for obtaining time-sensitive online public opinion data of food safety.(2)Research on topic discovery model of online public opinion on food safety.In order to improve the sparsity and smoothness of the topic discovery model,a non-negative Matrix Factorization(NMF)topic model based on Elastic Net was proposed,and the model was initialized by singular value decomposition.The nonlinear gauss-Seidel method is used to solve the model.Compared with Latent Dirichlet Allocation(LDA)topic discovery model,the results show that NMF topic discovery model has the advantages of short time and high quality topic acquisition.In addition,the topic evolution and hot topics are further analyzed according to the time of public opinion comments,the number of likes and replies.(3)Food safety network public opinion topic discovery system development.Based on Scrapy,Flask,Bootstrap and Mongo DB technologies,a food safety online public opinion topic discovery system with crawler control,data collection,data visualization and topic analysis was designed and implemented.The system realizes the visualization operation of crawling from the food safety network public opinion data to topic discovery based on Web,and is expected to be effectively applied to the monitoring of food safety network public opinion.