数据流中结构二叉树挖掘算法研究

针对传统数据流挖掘算法不能挖掘出频繁项之间的关系而且挖掘时间和空间复杂度高、准确度不高的问题，本文提出了一种数据流中结构二叉树挖掘算法（AMST）。该算法利用了二叉树结构的优势，将所处理事务数据库中的数据流转化成结构化二叉树，然后利用数据流矩阵对结构二叉树进行挖掘。整个过程只对事务数据库进行了一次扫描，大大提高了挖掘的效率。此外，算法还找出了具有层次关系的频繁子树。实验结果表明，AMST算法性能稳定，在时间复杂度和空间复杂度方面有很大的优越性，能够快速准确地对数据流进行挖掘。

Study on the algorithm for mining structural binary tree in data stream

Aiming at the problem that the traditional algorithms of data stream cannot mine the relationship between the frequent items, and the mining time and space complexity is high, while the accuracy is not high. An algorithm for mining structural binary tree in data stream (AMST) was proposed. The algorithm uses the advantage of binary tree, transforms the data stream into a structured binary tree, and then mines the binary tree with the data stream matrix. The algorithm greatly improves the efficiency of mining because it only scans the database once in the whole process. In addition, the algorithm also finds out frequent subtrees with hierarchical relations. Experiment results showed that the algorithm has stable performance and better advantage in time and space complexity, and can mine the data stream rapidly and accurately.