A systematic review of previous mobile payment studies revealed that studies mainly focused on consumer adoptions via conducting surveys, yet the low adoption of mobile payment services still exists. Despite a number of studies investigated barriers to mobile payment consumer adoptions, little explanation has been provided without taking ecosystem factors of mobile payment into account. Given the promising future of digital payments, we believe it is more relevant to study the actual usage of e-wallets via a novel approach. The objective of the study is to identify and categorize related themes of usages of e-wallets using big data analytics. With a large dataset of 18,149 user posts extracted from social media platforms, we apply the text mining method to analyze e-wallet users' behaviors. Our major findings are (i) whilst contradicting with user adoption factors (e.g., usefulness), users are attracted to use e-wallets to gain cashback and accumulate reward points; (ii) successful measures for e-wallet business models include a user-friendly interface, promotional campaigns, and customer service with real-time problem-solving; (iii) an intense competition between bank e-wallets and third party e-wallets is compounded with stricter government regulations; (iv) low rates of merchant adoption contribute to the lack of critical mass use of e-wallets. The big data analytics of actual usage of e-wallets produce more relevant and accurate understandings of the mobile payment mechanism. It reflects the complexity of human-computer interactions. This study establishes a predictive model of measuring successful e-wallet business and provides a holistic view of the mobile payment ecosystem with empirical evidence. The business must go beyond consumer adoption, learn actual user behavior, and take consumers’ pulse to achieve a sustainable business model. It is suggested that governments reform and upgrade the monitoring framework to accommodate the development of payment systems.

对以前的移动支付研究的系统回顾表明，研究主要通过进行调查来关注消费者的采用情况，但移动支付服务的采用率仍然很低。尽管有许多研究调查了移动支付消费者采用的障碍，但在没有考虑移动支付的生态系统因素的情况下，几乎没有提供任何解释。鉴于数字支付的未来前景广阔，我们认为通过一种新颖的方法来研究电子钱包的实际用途更为相关。该研究的目的是使用大数据分析来识别和分类电子钱包使用的相关主题。通过从社交媒体平台提取的18149个用户帖子的大型数据集，我们应用文本挖掘方法来分析电子钱包用户的行为。我们的主要发现是（i）在与用户采用因素（例如，实用性）相抵触的同时，吸引用户使用电子钱包获取现金返还并累积奖励积分；（ii）电子钱包业务模型的成功措施包括易于使用的界面，促销活动和具有实时解决问题能力的客户服务；（iii）银行电子钱包与第三方电子钱包之间的激烈竞争，加上更严格的政府法规；（iv）商家采用率低导致缺乏大量使用电子钱包的情况。电子钱包实际使用情况的大数据分析使人们对移动支付机制有了更相关，更准确的理解。它反映了人机交互的复杂性。这项研究建立了衡量成功的电子钱包业务的预测模型，并提供了具有经验证据的移动支付生态系统的整体视图。企业必须超越消费者的接受范围，了解实际的用户行为，并抓住消费者的脉搏以实现可持续的商业模式。建议政府改革和升级监测框架，以适应支付系统的发展。