查看文章

Volume, velocity, variety, veracity, and value are the main characteristics of big data; researchers consider them in the classification process. This study contemplates two of these characteristics (Data Volume and Veracity), as major attributes; the scale of data and accuracy proved to be issued in relation to varying boundaries. In the scenarios discussed by two methods, Tall array and MapReduce are used; as they were used to work with out-of-memory data. Tall array subdivides the data sets into small chunks that individually fit in memory, while MapReduce uses parallelization and distribution by enabling mapper function and reduce function respectively. Theoretical Model and Experimental simulation show that tall array method is more efficient compared to MapReduce as per F-Measure and Arithmetic Mean calculations; in tall array method, veracity is improved by 0.09 and 0.15 in respect to F-Mean and Arithmetic Mean, meanwhile volume is improved by 0.06 and 0.13.