Discovering closed and maximal embedded patterns from large tree data

We address the problem of summarizing embedded tree patterns extracted from large data trees. do so by defining and mining closed maximal unordered a single tree. design an frequent pattern algorithm extended with local closedness checking technique. This is called {\em closedEmbTM-prune} as it eagerly eliminates non-closed patterns. To mitigate generation intermediate patterns, we devise search space pruning rules to proactively detect prune branches in which not correspond The are accommodated into miner produce new algorithm, closedEmbTM-prune}, for all Our extensive experiments on synthetic real large-tree datasets demonstrate that, dense datasets, only generates complete set substantially smaller than that generated miner, but also runs much faster negligible overhead pruning.