Sub-frequent Patterns Mining Based on SFP-Tree

Resource distribution optimization and load balance of distributed P2P network can be described as mining sub-frequent patterns (SFPs) from query response transaction database. The response of a latter query for a file resource provided by P2P system will pick out one SFP from the SFPs of this file resource and returns a subset of selected SFP. To realize the SFPs mining this paper proposes the structure of SFP-tree along with relative mining algorithms. The main content includes: proposing the concept of sub-frequent pattern; proposing the SFP-tree along with frequency-ascending order header table FP-Tree (AFP-Tree) and conditional mix pattern tree (CMP-Tree); and proposing the SFPs mining algorithms base on SFP-Tree. The performance experiment shows the effectiveness and efficiency of SFP-Tree based mining algorithm.