| Since ancient times, reputation has been advocated unparalleled and pursuit perseverance by people. Only the person who has good reputation can stand between heaven and earth better, only the product which has good reputation can be favored by consumers. In addition, online reviews is in the period of rapid development. The major electricity supplier websites provide consumers a platform to publish online reviews. Consumers can be free to express their shopping and product using experience after buying related products, which can provide decision guidance for the subsequent potential consumers. This led to the development of online reputation which has been paid to more and more attention. This paper was based on online reviews and researched the product’s online reputation through the segmentation techniques and emotional semantic analysis to build mathematical model. The main research contents was product’s online reputation measurement based on online reviews and explored the effect of product’s online reputation on product sales. The main contents and conclusions are drawn as follows:(1) Based on product attribute thesaurus and evaluation emotional words gained by cooperation team through Chinese segmentation technology and emotional semantic analysis technology, this paper built product’s online reputation measurement model. On the basis of literature research, online reputation is defined as the evaluation of characteristics or nature of something or someone, which is expressed by UGC on the internet based on people’s own multi-dimensional perception to somebody or something at a time. This study abandoned the traditional way of reputation measurement—mental attitude scale method, and turned to the text analysis and emotional semantic analysis, built a mathematical model based on online reviews directly to measure the product’s online reputation.(2)This paper has selected related data of121kinds of cellphones in JD.com to verify the reliability of the product’s online reputation measurement model. In order to achieve this goal, This paper obtained online reputation of121kinds of cellphones based on measurement model, then made correlation analysis between the whole results and comprehensive ranking data of those cellphones provided by authority third-party websites, and the correlation coefficient is-0.320, and then selected two pairs of mutually competing products cellphones to make contrast analysis. The results showed that the online reputation measurement model is reliable.(3) Based on the online reputation of121kinds of mobile cellphones, combined with quantitative characteristics of online review and product attributes, this paper built several econometric models to explore the effect of online reputation on product sales. In order to eliminate multicollinearity between variables, we made the correlation analysis and found that there was a positive correlation between the average of star rating of online reviews and online reputation, the correlation coefficient is0.493. Therefore, we put the online reputation and the average of star rating as the independent variable respectively, along with the attributes of product itself, namely the control variables to construct two mathematical models to explore their effect on cellphone sales. We found that online reputation has a significant positive effect on product sales, but the comment star rating has no obvious effect on it. Then, based on the basis model, we introduced the number of reviews, poor assessment rate and the cross-term between online reputation and poor assessment rate one by one, the conclusions could be drawn as follows: online reputation and the number of reviews has a significant positive effect on cellphone sales, and the effect of online reputation on cellphone sales was regulated bay bad_rate; bad_rate has a significant negative effect on cellphone sales. This indicates that when we research the effect of online reputation on product sales, the number of reviews and bad_rate can not be ignored. |
PDF Full Text Request