It is very important to find whole frequent itemsets. To reduce the processing cost in real-time data mining, many approximate methods [10,37,39] have been proposed. This has recently also been a trend in the stream data environment. However, it is still important to find complete frequent itemsets for critical applications related to human life, critical facility management or analysis, and prediction for critical events. CanTree-GTree can find complete frequent itemsets.