The polarized problem of customers' emotional experiences in peer-to-peer accommodations has been exposed and needs urgent solution. However, a research gap remains in understanding the emotional heterogeneity. Therefore, this study explores customers' emotional heterogeneity and drivers for peer-to-peer accommodations using deep learning technologies and social network analysis. The results reveal that: For common causes, the environment is a core driver of customer emotions; While services are not necessary to positive emotions. For special causes, price value and location can stimulate customers' positive emotions, while their absence cannot affect negative emotions; The absence of household amenities cannot hinder the formation of positive emotions; While booking information triggers customers' negative emotions. This study clarifies complex demands of customer emotional experiences in sharing economy and provides a multiple emotional heterogeneity perspective for mining the emotional causes of peer-to-peer accommodation customers. Furthermore, it contributes to implications for improving customer emotional experience with differentiated strategies.

点对点住宿中客户情感体验的两极分化问题已经暴露出来，亟待解决。然而，在理解情绪异质性方面仍然存在研究空白。因此，本研究使用深度学习技术和社交网络分析探索客户的情绪异质性和点对点住宿的驱动因素。结果表明：对于常见的原因，环境是客户情绪的核心驱动力；虽然服务不是积极情绪所必需的。对于特殊原因，价格价值和位置可以激发顾客的积极情绪，而没有它们则不能影响消极情绪；缺乏家庭设施并不能阻碍积极情绪的形成；而预订信息会引发客户的负面情绪。本研究厘清了共享经济下顾客情感体验的复杂需求，为点对点住宿顾客的情感成因挖掘提供了多元情感异质性视角。此外，它有助于通过差异化策略改善客户情感体验。