OSSM: Ordered Sequence set mining for maximal length frequent sequences
The process of finding sequential rules is an indispensable in frequent sequence mining. Generally, in sequence mining algorithms, suitable methodologies like a bottom–up approach will be used for creating large sequences from tiny patterns. This paper proposed on an algorithm that uses a hybrid two-way (bottom-up and top-down) approach for mining maximal length sequences. The model proposed is opting to bottom-up approach called “Concurrent Edge Prevision and Rear Edge Pruning (CEG&REP)” for itemset mining and top-down approach for maximal length sequence mining. It also explains optimality of top-to-bottom approach in deriving maximal length sequences first and lessens the scanning of the dataset.
