Font Size: a A A

Laptop Based On Text Mining Online Review Research

Posted on:2021-02-28Degree:MasterType:Thesis
Country:ChinaCandidate:X D JiangFull Text:PDF
GTID:2439330620963701Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
With the rapid popularity of the Internet,people's lifestyles and working methods have undergone tremendous changes.With its advantages of high efficiency and portability,notebook computers have quickly become the first choice for long-term entertainment and work.The diversification of computer brands has made the market more fierce.This puts forward higher requirements for the performance of the notebook itself and after-sales service.How to improve the shortcomings on the basis of retaining the original advantages and improve their own competitiveness It seems very important.Therefore,this article takes the notebook computer as the research object,and takes the product improvement as the research purpose,mainly doing the following parts:The first part determines the specific research object and research method of the article.Based on the market share under the JD platform,this article selects Lenovo,Hewlett-Packard,and Dell as research objects,and selects popular products under the three major brands for text mining research,and provides reference for product improvement decisions based on their review information.After identifying the research object,read the relevant literature to determine the research method of the article.The second part contains the descriptive analysis of the parameters of each brand's laptop computer and the comparison of the hot comments of each brand's review content.First of all,a series of computer parameters of the top-selling laptops of Lenovo,Hewlett-Packard and Dell were crawled from the JD platform using the web crawler method.Then,the collected data will be visually displayed,that is,a descriptive statistical method will be used to analyze the general situation of the top three brands of hot products,so as to have a preliminary understanding.Finally,the hot words of the reviews of the three major brand consumers are presented in the form of word clouds,and a comparative analysis is made.The third part includes the establishment of computer attribute evaluation indicators and the analysis of Yiwang,a brand's poor reviews.First,the LDA topic model is used to classify the content ofthe review,and the review is divided into 12 topics according to the product attributes where the attribute words are located.Then calculate the satisfaction of each product attribute,user attention,variance of emotional value and the degree of urgent improvement of each product attribute,get the scores of the three major computer evaluation indicators,and compare and analyze it.The review of this attribute is further analyzed,and combined with the Semantic Web of the three major brand negative reviews to provide a reference for product improvement.Combining the above analysis,this article draws four conclusions:(1)the appearance difference is small,the brand characteristics are not obvious;(2)the quality of the three major brand notebook computers and to be strengthened;(3)from the merchants,logistics distribution to platform services are not in place;(4)Lack of personalized differences between brands and lack of brand competitiveness.At the same time,the following four suggestions are proposed in this paper:(1)Break the inherent style of appearance and innovate;(2)Strengthen the quality management in the computer production process;(3)Improve the service system from merchants,logistics and distribution to the platform;(4)Clarify and highlight brand characteristics.The research conclusions in this article enrich the existing theoretical results of the actual application of text mining,expand the application field of text mining,can provide manufacturers with decision-making basis for improving laptops,can provide reference opinions for logistics and distribution,and can provide new opportunities for businesses.The business ideas can provide a reference for the improvement of the service system of the Jingdong platform.
Keywords/Search Tags:Notebook computer, Descriptive statistics, LDA theme model, Evaluation indicators, Semantic network
PDF Full Text Request
Related items