Group-buying often fails even when there are satisfactory quantities as not enough consumers join in the required time, which can waste seller, purchaser, and platform operator time resources; therefore, the group buying features require further research. Over a 3 weeks period, around 700 million click-stream records from 1,061,770 users from a stable and continuous time series were allocated to groups of 5 min frequency, and a hybrid neural network model developed to simulate group-buying behavior in four experiments, from which it was found that adding to the cart and adding as a favorite were significant group-buying behavior features, and shopping depth was the main demographic feature, but age was not. Compared with previous ambiguous online consumer feature conclusions on gender, the results revealed that the commodity feature was the main determinant for gender feature significance. The college student feature was found to be a pseudo feature, and should connect with other fixed effects such as low income or education level. This paper is the first to construct an online dynamic group-buying community, which is a new type of social network and could provide a new perspective for social commerce research. A big data neural network-based method for examining group-buying community behavior over time is proposed that can offer novel insights to online vendors for the development of targeted marketing campaigns.

中文翻译：

---

基于高频时间序列仿真的在线动态团购社区分析

即使有足够的数量，团购通常也会失败，因为在规定的时间内没有足够的消费者加入，这会浪费卖方，购买者和平台运营商的时间资源；因此，团购功能需要进一步研究。在3周的时间内，将来自稳定和连续时间序列的1,061,770个用户的约7亿点击流记录分配给了频率为5分钟的组，并开发了一种混合神经网络模型来模拟四个实验中的组购买行为，分别来自结果发现，添加到购物车并添加为收藏夹是重要的团购行为特征，而购物深度是主要的人口统计特征，而年龄则不是。与之前关于性别的模糊在线消费者特征结论相比，结果表明，商品特征是性别特征意义的主要决定因素。发现大学生特征是伪特征，并且应该与其他固定影响（例如低收入或受教育程度）联系在一起。本文是第一个构建在线动态团购社区的平台，它是一种新型的社交网络，可以为社交商务研究提供新的视角。提出了一种基于大数据神经网络的随时间检查团购社区行为的方法，该方法可以为在线销售商开发目标营销活动提供新颖的见解。本文是第一个构建在线动态团购社区的平台，它是一种新型的社交网络，可以为社交商务研究提供新的视角。提出了一种基于大数据神经网络的随时间检查团购社区行为的方法，该方法可以为在线销售商提供针对目标营销活动的新颖见解。本文是第一个构建在线动态团购社区的平台，它是一种新型的社交网络，可以为社交商务研究提供新的视角。提出了一种基于大数据神经网络的随时间检查团购社区行为的方法，该方法可以为在线销售商开发目标营销活动提供新颖的见解。