Year: 2014
Journal of Information and Computing Science, Vol. 9 (2014), Iss. 2 : pp. 83–90
Abstract
FP-growth method is an efficient algorithm to mine frequent patterns, in spite of long or short frequent patterns. By using compact tree structure and partitioning-based, divide-and-conquer searching method, it reduces the search costs substantially. But just as the analysis in Algorithm, in the process of FP- tree construction, it is a strict serial computing process. Algorithm performance is related to the database size, the sum of frequent patterns in the database: ω. this is a serious bottleneck. People may think using distributed parallel computation technique or multi-CPU to solve this problem. But these methods apparently increase the costs for exchanging and combining control information, and the algorithm complexity is also greatly increased, cannot solve this problem efficiently. Even if adopting multi-CPU technique, raising the requirement of hardware, the performance improvement is still limited.
Journal Article Details
Publisher Name: Global Science Press
Language: English
DOI: https://doi.org/2024-JICS-22584
Journal of Information and Computing Science, Vol. 9 (2014), Iss. 2 : pp. 83–90
Published online: 2014-01
AMS Subject Headings:
Copyright: COPYRIGHT: © Global Science Press
Pages: 8