Discovering maximal subsequence patterns in sequence database

Mining sequential patterns in biological data has attracted a great deal of attention in the last couple of years. Biologists are interested in finding the frequent orderly arrangement of motifs that may be responsible for similar expression of a group of genes. The size of the output space can be greatly reduced if only the maximal frequent patterns are reported. In this paper we present maximal PrefixSpan algorithm which reports maximal frequent patterns in the sequence database. Experimental results on synthetic data shows that the size of the output space is greatly reduced when only maximal frequent patterns are reported.