Online Review Of Cold Chain Agricultural Products E-commerce Platform Based On Text Mining Emotional Analysis

The rapid development of e-commerce has led to the rapid rise of the cold chain agricultural products e-commerce platform,but there are still many problems in the development process.How to find and improve these problems and enhance consumers' willingness to purchase is the primary task.This article takes consumer reviews as the starting point and analyzes the sentiment.The analysis results can not only influence the purchasing decisions of other users,but also point the direction for the future development of the company.This paper takes the Huainan "vegetable basket project" cold chain agricultural product e-commerce platform as an example,extracts consumer reviews through crawlers,and adopts three methods of sentiment analysis.The specific research work includes the following three contents:First,based on Bi-LSTM,emotional classification of consumer reviews of cold chain agricultural products e-commerce platform.In order to help enterprises and consumers grasp the overall quality of goods,such methods are proposed from the perspective of textual statements.Firstly,the consumer comments are preprocessed,and based on word2vec training word vector,the relationship between the context of the review text is analyzed according to Bi-LSTM,and the attention mechanism is introduced to further optimize the model,and finally the sentiment classifier is used to evaluate the set.Conduct emotional classification.From the results,the consumer's overall evaluation of the "food basket project" e-commerce platform in Huainan is reflected,reflecting the consumer's satisfaction with the entire cold chain agricultural product e-commerce platform.Second,based on the LDA theme model,the online analysis of the cold chain agricultural product e-commerce platform is sentimental analysis.The approach examines consumer concerns from a potential topic perspective and extracts the most useful information for businesses and consumers.Firstly,according to the classification result of consumer reviews in Chapter 2,the emotional unit is extracted from the favorable and bad reviews by the dependency syntax relationship,and then the LDA theme model is used to cluster the favorable and bad reviews respectively,and the maximum value of the evaluation index SC value is determined.The number of topics,and finally the analysis of the results of the cluster,from the results can be understood to the consumer's focus,the most satisfactory platform and the most dissatisfied.Thirdly,based on the evaluation object extraction,the sentiment analysis of the online comment of the cold chain agricultural product e-commerce platform is carried out.For users,not only pay attention to the quality of the whole product,but also pay more attention to the behavior of various attribute qualities,and propose such methods.Firstly,based on the collocation rules of part of speech,evaluation words and evaluation objects,and using the word frequency and PMI method to filter and filter the evaluation objects,and using the text similarity to classify the evaluation objects into five categories,and finally calculate the rules for evaluating the polarity of the words,and calculate The emotional orientation of the "Dish Basket Project"e-commerce platform comment.From the results,we can see the satisfaction of consumers on the five attributes of Huainan's "vegetable basket project" e-commerce platform.From these three perspectives,this paper conducts an emotional analysis of the consumer reviews of the Huainan "vegetable basket project" cold-chain agricultural products e-commerce platform.From the perspective of consumers,timely understanding of consumer needs and evaluations,and taking targeted measures can solve The above problems are better to attract consumers and provide reference for other e-commerce platforms for cold chain agricultural products that are developing.Figure 13 table 14 reference 55...