给出了一种基于滑动窗口挖掘频繁序列算法。该算法给出了ε-近似序列集的定义,利用一种压缩的数据结构GSP-tree来存储和维护整个滑动窗口中各分区的近似序列集,并通过合并各分区的近似序列集来响应用户当前的查询请求。
A sliding window-based algorithm was proposed to mine frequent sequence. The definition of-approximate sequence set was given and a compressed data structure called "GSP-tree" was introduced to maintain the approximate sequence set of each partition in the whole sliding window.