Authors: T. Sheik Yousuf; M. Indra Devi
Addresses: Department of CSE, Mohamed Sathak Engineering College, Kilakarai, Tamil Nadu 623806, India ' Department of CSE, Kamaraj College of Engineering and Technology, Virudhunagar, Tamil Nadu 626001, India
Abstract: In data mining, the extraction of frequent patterns from large databases is still a challenging and difficult task due to the various drawbacks such as, high response time, communication cost to alleviates such issues, a new algorithm namely single scan distributed pattern mining algorithm (SSDPMA) is proposed in this paper for frequent mining. The frequent patterns are extracted in a single scan of the database. Then, it is split into multiple files, which will be shared to multiple virtual machines (VMs) to store and compute the weight for the distinct records. Then, the support, confidence and threshold values are estimated. If the limit is greater than the given data, the frequent data are mined by using the proposed SSDPMA algorithm. The experimental results evaluate the performance of the proposed system in terms of response time, message size, execution time, run time and memory usage.
Keywords: data mining; frequent pattern mining; single scan distributed pattern mining algorithm; SSDPMA; virtual machine; VM; file split algorithm; item sets; infrequent items; connect 4 dataset.
International Journal of Data Analysis Techniques and Strategies, 2019 Vol.11 No.1, pp.81 - 100
Available online: 29 Oct 2018 *Full-text access for editors Access for subscribers Purchase this article Comment on this article