移动社交应用成瘾因其严重的后果引发了广泛的社会关注。然而，目前大部分研究倾向于关注诊断判别和影响因素分析。虽然已有大量针对传统技术成瘾的文献研究，但本研究认为移动社交应用因其独特的使用特征产生了完全不同的使用环境与行为模式，因而存在显著不同的行为机制。为了全面研究移动社交应用成瘾这一现象，本课题主要研究以下三个问题：①如何科学有效地判别移动社交网络用户成瘾与否？②移动社交网络成瘾用户有哪些行为特征？③如何智能识别移动社交应用成瘾用户？在现有研究的基础上，本项目拟基于理性成瘾模型及移动应用间的替代互补性，结合客观数据和田野实验对用户是否成瘾进行判别，并对成瘾用户的行为特征进一步分析和挖掘。最后，利用机器学习的方法构建成瘾用户智能识别系统，以期及时发现并识别成瘾用户。研究结果有利于用户个人及社会更好地理解社交应用成瘾行为，掌握行为规律，并为干预和治疗移动社交应用成瘾提供依据。

Mobile social apps addiction has drawn extensive social attention due to serious and even fatal outcomes. Widespread concerns thus have inspired research in different disciplines. However, the focus so far tends to be on diagnostics and cognitive predictors. Although a significant line of literature exists in the area of personal-computer based technology addiction, the mechanism underpinning mobile social apps use differs significantly because the specific and unique characteristics have given rise to a fundamentally different usage context with new usage behavioral patterns. Motivated to comprehensively investigate this issue, this study attempts to explore the following questions: ① How should we effectively define mobile social apps users as addictive ones? ② What are the specific behavioral patterns of those addictive users? ③ How can we intelligently identify those addicted users of mobile social applications? Based on the rational addiction model, this study first defines addiction behavior of mobile social apps by incorporating the substitution/complemention effect among different mobile social apps. Using panel data analysis and field experiment, we then analyze and mine the specific behavior patterns of addictive users. We further use machine learning to build the intelligent recognition system to discover and identify those addictive users in time. The results of this study are helpful to users themselves and the society as a whole to better understand and master addiction behavior, and to provide suggestions for intervention and treatment of mobile social apps addiction.