基于在线评论情感分析模型的鲜果动态定价研究  

Fresh Fruit Dynamic Pricing Based on Sentiment Analysis of Online Reviews

作　　者：江亿平 张婷 夏争鸣 郑德俊[1] JIANG Yiping;ZHANG Ting;XIA Zhengming;ZHENG Dejun(Nanjing Agricultural University,Nanjing,China)

机构地区：[1]南京农业大学信息管理学院

出　　处：《管理学报》2022年第12期1837-1846,共10页Chinese Journal of Management

基　　金：国家自然科学基金资助项目(71803084);江苏省社会科学基金资助项目(21GLC003);江苏高校“青蓝工程”资助项目。

摘　　要：针对基于在线评论情感分析的鲜果动态定价问题,提出融合边缘采样和协同训练的在线评论情感分析方法,构建基于在线评论情感分析的鲜果动态定价模型,设计融合高斯回代的交替方向乘子法对模型进行求解,通过数值仿真和灵敏度分析,得出零售商最优定价决策。研究表明:正反馈情感的在线评论对鲜果销售有促进作用,零售商依据情感分析结果做出了提高鲜果价格的决策,获得了较其他定价策略更高的利润。此外,通过对比分析本研究提出的改进算法和传统的交替方向乘子法的迭代结果,发现交替方向乘子法与高斯回代有机结合具有更快的收敛性。This study addresses a problem of dynamic pricing of fresh fruit in terms of online review sentiment analysis. An online review sentiment analysis method that integrates edge sampling and collaborative training is proposed. And a dynamic pricing model of fresh fruit based on online review sentiment analysis is constructed. The proposed model is solved by an alternating direction multiplier method with Gaussian back substitution. Through numerical simulation and sensitivity analysis, the retailer’s optimal pricing decision is obtained. The results show that online reviews with positive feedback emotions can promote the sales of fresh fruit. Retailers make decisions to increase the price of fresh fruit based on the results of sentiment analysis and obtain higher profits than other pricing strategies. In addition, by comparing and analyzing the iterative results of the improved algorithm proposed in this study and the traditional alternating direction multiplier method, it is found that the organic combination of the alternating direction multiplier method and the Gaussian back substitution has faster convergence.

关 键 词：在线评论 情感分析 信息交叉 动态定价 融合高斯回代的交替方向乘子法 

分 类 号：C93[经济管理—管理学]