Some Observations of Sequential, Parallel and Distributed Association Rule Mining Algorithms

Garg, Rakhi; Mishra, Pragnyaban

doi:10.1109/iccae.2009.28

Cited by 8 publications

(7 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The algorithm switches back and forth between the two mining strategies. This approach is distinct to the prior related parallel works [17,23,24,25] which applied a single mining strategy. Figure 4 presents the overview of the parallel frequent pattern generation process.…”

Section: Frequent Pattern Generation 51 Parallel Frequent Pattern Gementioning

confidence: 97%

“…Several parallel FP-growth based methods have been proposed for shared memory multi-core systems. In the traditional approach, parallel processes cooperatively build a shared global FP-tree resulting in extensive use of costly synchronization locks to access each node of the tree [17]. A different approach called Tree Projection partitions the FP-tree into subsections with small portions shared among processes.…”

Section: Related Workmentioning

confidence: 99%

“…For example, Google deploys this task in their query recommendation system using MapReduce on a distributed-memory machine with thousands of cores [15]. Some surveys [16,17] have showed that few studies investigated on parallel frequent pattern mining algorithms for shared memory multi-core computers which underutilize the benefits of shared memory. None of previous parallel works took into consideration the data characteristics to improve the mining performance on different database types.…”

Section: Motivationmentioning

confidence: 99%

“…Majority of the existing parallel methods of mining frequent pattern have been proposed for distributed memory systems [15,16,17,18,19,20]. For the memory intensive problems like the one being investigated in this paper, efficient utilization of shared memory MIMD parallelism can significantly improve the overall performance [22].…”

Section: Related Workmentioning

confidence: 99%

“…Hence, it is essential to design and apply parallel computing techniques to speedup this mining task. Most existing works propose parallel solutions for distributed-memory systems [15,16,17,18,19,20]. For example, Google deploys this task in their query recommendation system using MapReduce on a distributed-memory machine with thousands of cores [15].…”

Section: Motivationmentioning

confidence: 99%

See 4 more Smart Citations

Novel parallel method for mining frequent patterns on multi-core shared memory systems

Alaghband

2013

Proceedings of the 2013 International Workshop on Data-Intensive Scalable Computing Systems

View full text Add to dashboard Cite

Frequent pattern mining is an important problem in data mining with many practical applications. Current parallel methods for mining frequent patterns unstably perform for different database types and under-utilize the benefits of multi-core shared memory machines. We present ShaFEM, a novel parallel frequent pattern mining method, to address these issues. Our method can dynamically adapt to the data characteristics to efficiently perform on both sparse and dense databases. Its parallel mining lock free approach minimizes the synchronization needs and maximizes the data independence to enhance the scalability. Its structure lends itself well for dynamic job scheduling resulting in well-balanced load on new multi-core shared memory architectures. We evaluate ShaFEM on a 12-core multi-socket server and find that our method runs 2.1 -5.8 times faster than the state-of-the-art parallel method. For some test cases, we have shown that ShaFEM saves 4.9 days and 12.8 hours of execution time over the compared method.

show abstract

Section: Frequent Pattern Generation 51 Parallel Frequent Pattern Gementioning

confidence: 97%

Section: Related Workmentioning

confidence: 99%

Section: Motivationmentioning

confidence: 99%