A Research On Reputation Analysis And Value Measurement Based On Users In Automotive Network Community

In the 21 st century,the world has entered an "Internet +" era,when online shopping has been favored by consumers in China.Besides,the development of e-commerce has made the network community a focus of many platforms or enterprises.In the network communities,the real experience of guests could be generally presented by grades,comments,and other information given by them.Additionally,with the assistance of previous online comments,other potential users can also identify the product quality,and business reputation.Meanwhile,both platforms and enterprises can also acquire useful information from comments,extract user wishes availably,measure the user value,classify users,and assist enterprises in identifying their own merits and demerits effectively,as well as adjusting business strategies and market strategies in time,aiming to make a profit.Therefore,it is of vital significance to use Information Technology to extract valuable information in considerable user comments.This thesis takes the Hong Qi brand from ATHM as an example,getting of the comments on the reputation from guests and the random comments from potential users.Based on the reputation comments from guests,first and foremost,this thesis perform the text analysis on the non-structured data,such as reputation comments,separating them then extracting the key words,counting the key words frequency,while processing visualized analysis and semantic analysis.What’s more,based on the structural data,such as the eight index scores given by guests and adopting regression analysis as well as other technical means,this thesis analyses the information characteristics of user comments and automotive properties,and conclude the ranking of automobile attribute.Last,due to the identity among the results obtained by above analysis methods and achievement of portray of vehicle and users’ characteristics,this thesis analyses the merits and demerits of Hong Qi brand,which provides theoretical and practical support for automobile enterprises’ service management and consumers’ decision.Based on the random comments of potential users,it first performs visual analysis,compared with the reputation comments of the previous car purchase users,and discover different users’ concerns.In order to assist the platform enterprises in measuring user values more accurately,and distinguishing different users comprehensively,considering the automobile as a bulk commodity,the traditional value measurement of the M value in the RFM model(the latest amount of consumption)is less useful.So,this thesis eliminates the indicator M in the traditional RFM model,reserves the index R(time interval)and F(frequency),and introduces three indicators of influence I,satisfaction S,and user emotion S,then constructs the RFISS user value measurement models,quantizing users value.While using the K-Means clustering algorithm to cluster,it could select the users with both high intention to consume and high brand loyalty,and provide the relevant marking strategies for auto manufacturers to expand market and market precisely.Finally,relying on the users’ characteristics,they are divided into 4 levels,the one of important value,common value,developing priority,and subordinate developing priority,and the corresponding marking management is raised,aiming to ensure the user value is maximized.