| It has already received universal attention from academics and industry to measure celebrity endorsement value.The scholar Huang heshui(2011)has proposed a measure model to evaluate celebrity endorsement value based on network.But this model also has some disadvantages on the measurement and calculation.In order to correct the shortcomings of the model,1106806 pieces of news of 536 celebrities in 2015 are downloaded and these headlines are analyzed by the method of contentanalysis in this study.This study focus on building a set of stable and effective method to measure the celebrity endorsement value.To achieve this goal,this study has been broken down into three little studies,respectively about the classification,retrieval,and calculation.Research results are as follows:(1)The study one has optimized the measure dimensions of Huang heshui(2011)scholar's measurement model about celebrity endorsements value.And adds the dimension of celebrity scandal based on the model.Celebrity scandal contains breach behavior,illegal behavior of two core categories,which includes ten segmentation indexes.This study also concludes the segmentation indexes of four core categories on the original model about attractiveness image,promotion image,trustworthiness image and identity image.Such as attractiveness image can be divided into two indexes,like external charm and popularity degree;Trustworthiness image can be made of credibility,professionalism,public image,personal honor and so on.The reliability test value of six core categories are above 0.90.This coding tool can be used for the measurement of celebrities' news.(2)According to the results of study one,the study two has summarized the key words of celebrity image and celebrity scandal,and formed a set of keywords retrieval tool based on news headlines.All news about core categories can be detected by 88 to 313 keywords.Among them,there are three set of keywords about attractiveness image and trustworthiness image,group B and C should be used in combination.Other main categories only have one set of keywords,can directly retrieve out related news.The concordance rate between the keywords-based retrieval and artificial coding results is over 98%.And more than 93%of the correct news headlines can be retrieved by using these keywords,thus improve the coding efficiency.Moreover,the study two sums up some relatively stable keywords about all core categories through focus group interview and sampling data validation in nearly three years.Such as men' attractiveness image can be described as "handsome" "cool""sexy" and other words."Beauty" "beautiful" "sexy" or "sweet" are relatively stable words for women.Hot words on the internet such as "god" "goddess" and "fresh meat",though,don't stay hot for long,but can be seen as important keywords to evaluate the attractiveness image in recent years.(3)Based on the previous two studies and combined the theory of Huang heshui's model(2011),the study three optimizes the calculation method to evaluate celebrity endorsement value.By using this algorithm,the celebrity endorsement value index and the ranking of each dimension in 2015 are obtained.Among them,we find that the most valuable top ten female celebrities are,in order,Fan bingbing,Yang mi,Zhao liying,Tang yan,Zhao wei,Sun li,Zheng shuang,Angelababy,Liu yifei,Deng ziqi.And male celebrities are Jay zhou,Huang xiaoming,Li chen,Lu han,Li yifeng,Wu yifan,Hu ge,Deng chao,Jackie chan,Xie tingfeng.The ranking results through the comparision of "the Forbes list of celebrity" as well as other lists,is considered to be reasonable and effective. |
PDF Full Text Request