Full Perfect Extension Pruning for Frequent Graph Mining

Mining graph databases for frequent subgraphs has recently developed into an area of intensive research. Its main goals are to reduce the execution time of the existing basic algorithms and to enhance their capability to find meaningful graph fragments. Here we present a method to achieve the former, namely an improvement of what we called "perfect extension pruning" in an earlier paper (Borgelt, 2004). With it the number of generated fragments and visited search tree nodes can be reduced, thus accelerating the search