使用序列模式精简基挖掘序列模式

传统的序列模式挖掘方法在挖掘由短的频繁序列模式组成的数据库时有良好的性能．但在挖掘长的序列模式或支持度阈值很低时，这些方法可能遇到固有的困难，因为产生的频繁序列模式的数量经常太大．在许多情况下，用户可能只需要那些覆盖许多短模式的长模式．此外，在很多应用中，只要得到产生的频繁序列模式的近似支持度就已足够，而不需要它们的精确支持度．介绍了能将误差控制在确定范围内的频繁序列模式精简基的概念，并开发了一个挖掘这种序列模式精简基的算法．实验结果显示计算频繁序列模式精简基是很有前途的．

Using Condensed Sequential Pattern Bases to Mine Sequential Patterns

Conventional sequential pattern mining methods have good performance in databases consisting of short frequent sequences. When mining long frequent sequences, or when using very low support thresholds, such algorithms may meet inherent difficulties since the number of frequent sequential patterns generated is often too large. In many situations, a user may just want the long patterns that cover many short ones. Moreover, in many applications it is sufficient to generate only frequent sequential patterns with support frequency in close-enough approximation instead of in full precision. This paper introduces the concept of condensed frequent sequential pattern base with guaranteed maximal error bound, and an algorithm is developed to mine such a condensed sequential pattern base. Experimental results show that computing condensed frequent sequential pattern base is promising.