A Depth-First Search Approach for Mining Proportional Fault-Tolerant Frequent Patterns Efficiently in Large Database

Mining of frequent patterns in databases has been studied for several years. However, real-world databases contain noise and frequent pattern mining which extracts patterns that are absolutely matched is not enough. Therefore, a research field called fault-tolerant frequent pattern (FT-pattern) mining is proposed to deal with this problem. In this paper, we consider the problem of mining proportional FT-patterns. That is, the number of faults tolerable in a pattern is proportional to the length of the pattern. To reduce the disk I/O times, a depth-first mining approach is proposed to mine proportional FT-patterns efficiently in large database. Moreover, a set of experiments is performed to show the advantage of the approach. Experimental results indicate that the proposed algorithm outperforms the other existing approach when the database size is large.