摘要
采用定量更新滑动窗口策略挖掘数据流频繁闭项集,提出TW-CFI(Time Windwos-Closed Frequent Itemsets)算法更好的解决了数据流频繁模式挖掘问题,挖掘过程中采用一种CFIT(Closed Frequent Itemsets Tree)存储结构,不仅包含当前数据流中的频繁闭项集,还通过时间窗口保存频繁闭项集的历史频繁计数。
An efficient algorithm based on ration sliding window strategy is proposed, which name is Time Windows-Closed Frequent Itemsets (TW-CFI). Utilizing the algorithm, we can significantly discover frequent closed itemsets in data stream. A new data structure called Closed Frequent Itemsets Tree (CFIT) is designed, the structure not only includes the current frequent closed itemsets, but preserves the history counts of old frequent closed itemsets in virtue of time windows.
出处
《软件》
2012年第12期101-102,共2页
Software
关键词
数据流挖掘
频繁闭项集
滑动窗口
时间窗口
mining data steam, closed frequent itemsets, sliding window, time window