BIG DATA SOURCES AND DATA MINING

Abstract

Data mining is vial in every way in every industry starting from small vendors to the large MNCs. Big data is comprises large-volume, complex, growing data sets with multiple, heterogeneous, autonomous sources. The growth of networking, data storage and data collection capacity are growing in leaps and bounds, this necessitates the need of Big data in all domains. Big data is expanding in all engineering and science domains, including physical, biological and biomedical sciences. This paper presents a HACK theorem that proposes a Big data processing model from the data mining perspective. Also it characterizes the features of Big data revolution. This model is based on data-driven character of information which also involves a demand-driven aggregation of information sources, mining and analysis, user interest modeling, and security and privacy considerations. Throughout this paper we focus on analyzing the challenging issues in the data-driven model and also in the revolution phase of Big data.