基于AFOPT-tree的最大频繁项集挖掘

Mining of maximal frequent item sets based on AFOPT-tree

在最大频繁项集的挖掘过程中,尤其在数据规模庞大并且最小支持度较小的情况下,超集检测成为算法运行的主要时间消耗,提出最大频繁项集算法A-MFI,其通过优化基于投影的超集检测机制有效地减少了超集检测的时间。另外,将事务数据库数据映射至一种压缩的AFOPT-tree结构,该结构结合自顶向下的遍历策略,具有更小的时间开销。

During the process of mining maximal frequent item sets, when the data size is large and minimum support is little, superset checking is the main time-consuming in the mining algorithm. In this paper, a new algorithm A-MFI is proposed, by improving the superset checking method based on projection of the maximal frequent item sets, it efficiently reduces the cost of superset checking. In addition, the databases are projected onto a compact structure called AFOPT-tree which costs few time in a topdown traversal strategy.