An approach for the implementation of semantic Big Data Analytics in the Social Business Intelligence process on distributed environments (Cloud computing)

Abstract

Managing and extracting useful knowledge from social media sources is a challenge. It has attracted a lot of attention from universities and industry. To meet this challenge, semantic analysis of textual data is the subject matter. Today, with the connection present everywhere and at any time, considerable data is born. These data or data become a key player for understanding, analyzing, anticipating and solving major economic, political, social and scientific problems. Data also changes our working procedures, our cultural environment, even restructuring our way of thinking. And just as the scientific, managerial and financial world is interested in Big Data, a new discipline is growing: Fast Data. In addition to the salient volume of data; another variant becomes decisive, the ability to efficiently process data in all their diversity, transforming it into knowledge by providing the right information to the right person at the right time, or even using it to predict the future. The exploitation of Big Data requires the proposition of new adapted mathematical and IT approaches but also a reengineering of managerial approaches for the control of the informational environment of a public or private organization. While basing itself on a strategic information management approach such as Economic Intelligence (EI). The latter combines and encompasses Business Intelligence techniques for internal data management and business intelligence techniques for monitoring and controlling external information flows. However, Big Data, as a boundless source of information for EI, has upset the traditional EI process, which requires a reengineering of the EI approach. My research works perfectly in this context characterized by an uncertain and unpredictable environment. We ask to propose an ontology-based, service-oriented, agile and scalable Social Business Intelligence approach to extract the semantics of textual data and define the domain of massive data. In other words, we semantically analyze social data at two levels, namely the level of the entity and the level of the domain.