海量高维数据下的频繁项目集挖掘算法研究

摘要

Frequent Item sets mining is one of the most important directions of research in data mining area. In recent years, the informa-tion technology, represented by the internet, is expediting the course of global digitisation. While bringing us with convenience, it also pro-duces the data in enormous size. Moreover, these data have very high dimensions. Therefore, focusing on the large scale data at present, our research is on the raining algorithm for frequent item sets with massive and high dimensional data. In this paper, we propose a mining algo-rithm for frequent item sets which is effective, load balanced and good in scalability, and at the same time design a distributed algorithm based on MapReduce programming model. Experimental results show that our algorithm has noticeable advance than traditional algorithm in both time complexity and space complexity.%频繁项目集的挖掘是数据挖掘领域最重要的研究方向之一.近年来,以互联网为代表的信息技术加速着全球的数字化进程,在给人们生活带来便利的同时也产生了规模非常庞大的数据,而且这些数据的维数非常高.因此,针对目前的大规模数据,主要研究海量高维数据的频繁项目集挖掘算法.提出了高效的、负载均衡的、扩展性良好的频繁项目集挖掘算法,同时设计了基于MapReduce编程模型的分布式算法.实验结果显示,该算法在时间复杂度和空间复杂度上相比传统算法都有明显的提升.