An Analysis On The Affective Tendency Of Commodity Online Comment On Intelligent Speaker In Jingdong Mall

With the popularization of the concept of Internet of things and high point of the development of artificial intelligence,intelligent home appliances have entered People’s Daily life.Intelligent speaker is one of the most competitive products in the intelligent home appliance industry.All along,smart speaker is regarded as the future home intelligent control terminal,which can not only conveniently control smart home devices,but also integrate smart life services.However,at present,consumers have different opinions on the smart speaker in the market,which indicates that this emerging product still has many deficiencies while bringing convenience to people’s life,and its functions and performance cannot meet the numerous needs of consumers for the time being.At the same time,online shopping has become common part of daliylife,and online comments have become an important way to reflect people’s emotional attitude towards online shopping.By analyzing the online evaluation of online shoppers,we can effectively obtain important information such as the advantages and disadvantages of products and consumers’ demands,and provide Suggestions for the improvement of products and the marketing of merchants.Therefore,if the smart speaker wants to survive in the market environment with fierce competition,it is particularly important to study the emotional tendency of its online comments,accurately analyze and explore its shortcomings in design and sales.In recent years,with the rapid development of Internet technology and e-commerce platforms,online shopping has become more common.After online shopping,people can comment on products and merchants according to the goods they receive and the attitude of customer service.Online comments have become an important way to reflect people’s emotional attitude towards online shopping.By analyzing the online evaluation of online shoppers,we can effectively obtain important information such as the advantages and disadvantages of products and consumers’ demands,and provide Suggestions for the improvement of products and the marketing of merchants.This paper aims to find out the advantages and disadvantages of smart speakers in different attributes by analyzing the emotional tendency of online comments of online users of smart speakers.In this paper,through collecting Jingdong mall intelligent speakers online reviews data in 25372,the first by Chinese word segmentation,to stop words,part-of-speech tagging,and word vectorization handle collection of text preprocessing methods,such as intelligent speaker comment on data sets,and then the text analysis in the field of traditional emotional dictionary matching method and the SVM in the field of machine learning and neural network model is compared and analyzed in the analysis of Chinese text classification of emotional effect better model,at last the text classification model combined with theme LDA model,mining online comment on the theme of the property,The TextCNN model was used to conduct emotional evaluation for each subject attribute,analyze the emotional tendency of the intelligent speaker in each attribute,and propose specific Suggestions and improvement directions for each attribute of the intelligent speaker.The results show that,first,using the improved borderline-smote algorithm can effectively solve the non-equilibrium problem of the dataset.Second,the TextCNN model has better classification accuracy,recall rate,F1 score and other indicators in the smart speaker comment data set than dictionary matching and SVM model.Thirdly,the function of the intelligent speaker is complete,which perfectly reflects the integration concept of home appliances.The sound quality of the speaker is clear and has been recognized by the majority of consumers.It is suitable for the elderly and children to use and is a suitable gift choice.Based on the research results of this paper,it is suggested that smart speakers continue to expand their advantages in function types,applicable population,sound quality and other attributes of the speakers,and improve their endurance capacity,library resources,sensitivity and other attributes,so that the functions and performance of smart speakers can better meet the needs of consumers.