中文摘要：

数据挖掘是当今研究的一个热点，传感器实时收集大量的数据，将数据收集与数据挖掘技术结合起来，是现代数据处理技术发展的重要趋势。频繁模式挖掘是数据挖掘中的核心问题，本文针对数据库发生变化时频繁模式挖掘中普遍存在的重复扫描、遍历和计算问题，提出了频繁模式的增量维护算法IM-FPM。该算法充分利用已有挖掘结果来提高效率但又完全独立于上次采用的挖掘方法，并且只需对原始数据库进行一次扫描。实验结果表明，该算法能有效地解决数据库发生变化时的频繁模式增量维护问题。

英文摘要：

Data mining is still an important subject nowadays. Along with a mass of data collected by sensors, the data collecting combined with data mining technique becomes a main trend. Mining frequent pattern has been studied popularly in data mining research. However, very little work has been done on maintenance of mined frequent patterns. Costly and repeated database scans and structure travels were done due to not maintaining the discovered frequent patterns. A new algorithm, IM-FPM, is proposed to improve the efficiency of incremental maintenance problem of frequent pattern mining when a new database is inserted and the minimum support is not changed. The IMFPM algorithm uses effectively the patterns mined to improve the efficiency, but it is independent of the method adopted before. It only needs to scan the original database once. The performance study shows that the algorithm is efficient for incremental maintenance of frequent patterns mining.