An Efficient Access Reduction Scheme of Big Data Based on Total Probability Theory

Big data is being widely used in various fields and the accuracy and calculation cost regarding the search results of big data are being researched constantly. In this paper, a big data access reduction scheme based on total probability theory is proposed to improve the accuracy and minimize calculation cost of big data search. The proposed scheme uses the reduction approach of divide-and-conquer; that is, it distinguishes all the attributes of data so as to minimize the data. Also, to improve the efficiency of data access, the proposed scheme assigns the attribute information in accordance with the properties of big data access to minimize the required amount of information to classify the information in the big data group into tuple based on the probability values, in order to apply the least randomness within the big data group. In particular, the proposed scheme aims to improve data access compared to the existing methods by connecting the probability values among the data to access the divided data more easily. The performance evaluation results show that compared to the existing method, the proposed scheme improved accuracy by 7.1%, decreased data storage space by 3.8%, and shortened the process time by 11.1%.