基于BP神经网络模型的用户问答对服装产品月销量影响
2021-07-09张炎亮赵蓓
张炎亮 赵蓓



摘要: 用户问答作为产品口碑的新型定性化体现,一经推出迅速成为各大电商平台的热点。文章以网购平台中服装产品的用户问答为切入点,构建BP神经网络分析产品月销量的关键影响因素,对电商平台管理、店铺经营具有实践意义。借助分析工具对用户问答数据内容进行挖掘,构建用户问答效价词典,并建立用户问答相关变量与产品月销量的BP神经网络模型进行实证分析。神经网络仿真结果表明,各相关变量对月销量存在不同程度的影响,用户问答数量正向影响月销量,用户问答类型中的关键影响因素为属性型回答。
关键词: 用户问答;在线评论;文本分析;情感分析;BP神經网络;产品月销量
中图分类号: TS941.1
文献标志码: A
文章编号: 1001-7003(2021)05-0070-06
引用页码: 051111
DOI: 10.3969/j.issn.1001-7003.2021.05.011(篇序)
Abstract: As a new qualitative embodiment of product reputation, customers questions and answers(Q & As) has quickly become a hot spot on major e-commerce platforms once launched. This paper takes the customers Q & As in the online shopping platform as the entry point to build a BP neural network to discuss the key influence on the sales volume. By using the text analysis tools to mine the content of customers Q & As data, construct user Q & As valence dictionary, and establish the BP neural network model of cutomers Q & As related variables and monthly product sales for empirical analysis. The results of neural network simulation show that the relevant variables have different degrees of influence on the monthly sales volume, the number of user questions positively influences the monthly sales volume, and the key influencing factor of user questions type is attribute answers.
Key words: customers Q & As; online reviews; text analysis; sentiment analysis; BP neural network; monthly sales of products
随着互联网的普及和飞速发展,网络购物成为大众主流的消费方式。传统网络购物过程中,消费者通常通过在线评论来预测和判断商品的质量与口碑[1-3]。然而商品口碑包含多个方面,如品牌、店铺信誉等都对消费者的购买行为产生重要的影响[4]。研究发现,口碑丰富能够更好迎合顾客需求的多样性,因此网络平台也推出了更多新的功能板块来完善客服体验[5],如淘宝网在商品详情页所推出的“问大家”和京东商城的“问答”板块,通过用户问答(Customer Q & As)建立起一个用户直接传播口碑的平台[6]。通过该板块,潜在消费者可以就产品感兴趣的方面向已购买产品的消费者发出提问,淘宝后台邀请买家进行回答,消费者根据买家回答再进行购买决策。……
