动态数据库中增量Top-k高效用模式挖掘算法

高效用模式的挖掘需要设定一个合适的阈值，而阈值设定对用户来说并非易事，阈值过小导致产生大量低效用模式，阈值过大可能导致无高效用模式生成。因而Top-k高效用模式挖掘方法被提出，k指效用值前k大的模式。并且大量的高效用挖掘研究仅针对静态数据库，但在实际应用中常常会遇到新事务的加入的情况。针对以上问题，提出了增量的Top-k高效用挖掘算法TOPK-HUP-INS。算法通过四个有效的策略，在增量数据的情况下，有效地挖掘用户所需数量的高效用模式。通过在不同数据集上的对比实验表明TOPK-HUP-INS算法在时空性能上表现优异。

An incremental Top-k high utility pattern mining algorithm in dynamic database

Setting an appropriate threshold is essential to high utility pattern mining, however it is a difficult task for users. If the threshold is too low, a large number of low utility patterns are generated; if the threshold is too high, no high utility patterns may be generated. According, Top-k high utility pattern mining is proposed, in which k is the number of highest utility patterns. Most studies on high utility mining are only designed for static database, yet in real-world applications some new transactions inserted in original database. To address the above issues, an incremental Top-k high utility pattern mining algorithm is proposed, named TOPK-HUP-INS. The algorithm adopts four effective strategies to mine the high utility patterns that the number is user-specified in the case of incremental data. Comparing the experimental results on different datasets show that TOPK-HUP-INS performs well in terms of time and space.