Mining Sequential Patterns for interval based events by applying multiple constraints

Abstract

Sequential pattern mining finds the frequent subsequences or patterns from the given sequences. TPrefixSpan algorithm finds the relevant frequent patterns from the given sequential patterns formed using interval based events. In our proposed work, we add multiple constraints like item, length and aggregate to the interval based TPrefixSpan algorithm. By adding these constraints the efficiency and effectiveness of the algorithm improves. The proposed constraint based algorithm CTPrefixSpan has been applied to synthetic medical dataset. The algorithm can be applied for stock market analysis, DNA sequences analysis etc.