Data Mining of Movement and Action Logs of MMORPG Players

研究成果涵盖了三个提高网络游戏服务质量的数据挖掘主题，即：1.基于动作和/或移动日志的用户聚类；2.基于项目日志的不端用户检测；3.基于游戏重访模式的退休用户预测。除了在线游戏，这些结果对于在无处不在的IT环境下提供个性化服务的应用程序也具有很高的潜力。

Research findings cover three data mining topics for improving the service quality of online games, i.e., 1. user clustering based on action and/or movement log, 2. detection of misconduct users based on item log, and 3. prediction of retiring users based on game revisit patterns. In addition to online games, these results also have high potential for applications that provide personalized services under ubiquitous IT environment.