An improved approach for sequential utility pattern mining

In this paper, we propose an efficient projection-based algorithm to discover high sequential utility patterns from quantitative sequence databases. An effective pruning strategy in the proposed algorithm is designed to tighten upper-bounds for subsequences in mining. By using the strategy, a large number of unpromising subsequences could be pruned to improve execution efficiency. Finally, the experimental results on synthetic datasets show the proposed algorithm outperforms the previously proposed algorithm under different parameter settings.