An Efficient Approach for Mining Frequent Patterns Based on Traversing a Frequent Pattern Tree

Mining frequent patterns is an important task for knowledge discovery, which discovers the groups of items appearing always together excess of a user specified threshold. A famous algorithm for mining frequent patterns is FP-Growth which constructs a structure called FP-tree and recursively mines frequent patterns from this structure by building conditional FP-trees. However, It is costly to recursively construct conditional FP-trees. In order to decrease the usage of memory space and speed up the mining process, we propose an efficient approach for mining frequent patterns. Our approach only needs to construct a FP-tree and traverse each subtree of the FP-tree to generate all the frequent patterns for an item without constructing any other subtrees. Since there is no extra trees constructed and only a subtree needs to be traversed to generate frequent patterns for an item, our approach is much more efficient than FP-Growth algorithm. The experimental results also show that our approach significantly outperforms FP-Growth algorithm.