Application Research Of Text Sentiment Analysis In Commodity Evaluation

With the rapid development and popularization of Internet technology,online shopping has brought great convenience to people's life,becoming an important form of shopping.In order to enhance the customer's participation enthusiasm,the e-commerce platform allows customers to evaluate the product after receiving it.These comments reflect the customer's emotional attitude towards the function or performance of the product directly.Accordingly,the research on user's sentiment orientation on the products in the review text is very significant.Nevertheless,the data of online review is huge,as well as semi-structured and unstructured,and the useless comments is more.Therefore,how to acquire and analyze the goods comment text rapidly becomes the key issue in the study.To solve the above problem,based on the SparkR platform,this paper presents the application of text emotion analysis technology to explore the feature attribute information in commodity comment.At the same time,since the advent of artificial intelligence and the change of people's lifestyle,the paper selects objects of the top five intelligent refrigerator comments of jingdong mall.The evaluation of the attributes in the market is obtained through the analysis of the emotional polarity of the intelligent refrigerator in the review text.First of all,this paper uses the built SparkR platform to get the comment data of the smart fridge through the crawler,and stores it into the MySQL database according to the brand as well as carry on the data preprocessing operation to the comment corpus.On the basis of the preliminary analysis of the commentary text into positive emotion and negative emotion,we select the commentary with obvious emotion classification as the training set,and use K-means clustering algorithm to classify again to improve the accuracy of classification.Then,after the positive and negative classification of the text after the word segmentation,part of speech,subject analysis and syntax analysis.Finally,based on syntactic relations,the classification of special words and emotional words are extracted,and the emotional extremum of the characteristics of the intelligent refrigerator is calculated.Five brands of research results show that compared with the industry's top five intelligent refrigerator performance,Haier gets a higher score in the appearance,noise control,heating control,after-sales service and have a market competitive advantage,but there appears the deficiencies of preservation,Cost-effective,and lack of intelligence as well;The preservation,energy consumption,fever and other aspects of Meiling are at the mid-level of the five refrigerators,but the design and noise control design is poor;Siemens smart refrigerator has a higher score in the design of intelligence,compressor,noise control and technological level,but its price is on the high side;Samsung gets a better market evaluation in the freezing speed and compressor performance design,holding advantage at market competitiveness.However,the noise,price,energy consumption and after-sales service of Samsung are on the poor performance.Samsung receives the worst evaluation of the five kinds of refrigerators in the market;Midea has a high score in smartness,appearance,space,quality,cost performance,performance and other aspects.So it has a strong competitive advantage in the industry leading position.But it gets lower scores in energy consumption,fever,style design,internal design,packaging,which needs to be strengthened in the future design to enhance its market competitiveness.