Analysis Of Food Safety And Public Opinion Data Based On LDA Theme Model

With the rapid development of the Internet age,people can obtain a lot of information through various public platforms on the Internet.The huge amount of information has also become a development trend of the big data network in the current era.The issue of food safety is an important topic that deserves attention.Therefore,it is very meaningful to find important issues related to food safety from a large amount of information.Latent Dirichlet Allocation(LDA),as a potential semantic topic model,can realize the association between subject vocabularies between the semantics of the same words,so as to realize the analysis and clustering of the implicit vocabulary in the text.The document or the subject of each document in the document set is listed in the form of probability distribution,which is more suitable for us to analyze the public opinion data analysis of food safety issues.The main work of this paper is as follows:(1)The first is to get data.Here is the crawler technology.There are several crawling strategies for web crawler technology,so for each strategy and rule comparison,choose the one that is best suited for this topic.(2)Processing the acquired data,the data processing is to ensure the quality of the data,so this process is also very important,is divided into two steps,the first step is to remove the data,is to remove useless information;second The step is to process the data word segmentation.Through the research on the technology of word segmentation,it is found that the most commonly used is the word segmentation and the NLPIR word segmentation technology of the Chinese Academy of Sciences.This paper selects the word segmentation to operate.(3)LDA is used for model analysis of the processed data.Because LDA is an unsupervised learning technology,the LDA is tagged as a semi-supervised model for comparative analysis,including the model in LDA.The parameters are solved using Gibbs Sampling.Finally,through analysis and comparison of several models with the LDA topic model and the semi-supervised LDA topic model,it can be concluded that it is very helpful in the analysis of public opinion data for food safety.Given the importance of food safety,this research is very Valuable,based on this theory,it is possible to carry out public opinion analysis on food safety issues.