Toward a New Approach for Real-Time and Semantic Big Data Integration

Abstract

The exponential growth of data derived from the Internet exceeds the social networks plateforms such as Facebook, Twitter and so on. This data includes a wide range of sensor data from a variety of sources, such as IoT devices, satellite data and so on. This massive amount of data led to the emergence of ”Big Data” concept. This new trend has had a considerable impact, particularly in the ffeld of decision support systems. In particular, these impacts are mainly observed in the extraction, transformation and loading processes within business intelligence systems. In this context, three main challenges emerge: dealing with large quantities of data, various types of data and rapid data generation. In this paper, we present and discuss the state of the art of research focused on ETL processes while addressing the challenges of big data. Our study aims to determine the degree to which these works take into account the characteristics of big data in their approaches. Finally, we provide an overview of a new approach to ETL processes, called BRS-ETL, which supports heterogeneous and streaming data.