一种新的增量式关联规则数据挖掘方法研究

Study on a new method of incremental association rule data mining

本文介绍了一种实用的增量式关联规则数据挖掘方法,它映射数据挖掘空间到图像空间,并利用数字图像显示及存储的特点,结合增量事务的挖掘过程进行挖掘。该方法在一定程度上提高了关联规则数据挖掘的实时性,对于数据集及支持度的变化具有很好的适应性。通过铝电解现场采集数据挖掘实践证明,此增量式挖掘方法简单有效,具有很强的实用性;同时为提高铝电解生产的管理水平,提供了一种对工人操作可靠性、系统控制稳定性衡量的客观评价手段。

This paper describes a practical method for incremental association rule data mining. Mapping data mining space into image space, the method combines mining process of incremental transactions and mines data using the display and storage features of digital image. The method improves the real-time quality of data mining to a certain extent and gets good adaptability on the change of dataset and support. Experiments on the real data of aluminum reduction production prove that the proposed method is easy, effective and has high practicability. Moreover, in terms of improving the management level of aluminum reduction production, the proposed method provides an objective judgment measure to evaluate the worker operation reliability and system control stability.