期刊名称：International Journal of Advances in Soft Computing and Its Applications
印刷版ISSN：2074-8523
出版年度：2018
卷号：10
期号：1
页码：55
出版社：International Center for Scientific Research and Studies
摘要：During recent years, scientists have endeavored to meet the challengingnecessities of creating top notch programming applications. Building ventureclass arrangements like a Big Data Analytics Platform includes testingenormous measures of clinical trial information from different investigations,subjects, and installed gadgets. Handling and putting away terabytes orpetabytes of information may take days or weeks to finish. Utilizing an extensiveinformational collection amid programming advancement and testing,postpones the constant incorporation and conveyance endeavors. A novelmethod has been proposed to reduce the input data set for such big dataapplications without compromising on quality of results by creating a smallerrepresentative sample out of given big data datasets and using thatrepresentative sample to drive the application further. The proposed approachmakes use of Pairwise Test Case Generation methodology to identify hediminished arrangement of datasets ensuring that each pair of input parametershave been enclosed by at least one input data tuple. Although using pairwisemethodology might not be exhaustive, but it is found to be very useful because itcan significantly reduce the input dataset and can still cover hypotheticallydifficult relations within different input domain parameters. The paperdescribes the proposed hybrid pairwise test generation approach based uponCuckoo Search (CS) and Genetic Algorithms (GA) and applies the approach onbig data dataset to come up with representative data sets which are muchreduced as compared to original ones. The big data datasets taken as input toprove the effectiveness of proposed approach are the test cases intended to beused for defect finding. The quality of newly generated representative datasets with proposed approach is also evaluated against the original big sized big data datasets.
