Design process of a Big Data architecture for analysis of large volumes of data and information

Abstract

The objective of this article is to present a design of a Big Data Architecture for financial institutions, which allows the analysis of large volumes of data and information and promotes better decision making in less time. For this purpose, several scientific methods and techniques were used to allow the analysis, information extraction and validation of the proposed architecture.  This is divided into three parts: obtaining data in a structured and unstructured manner from different sources, processing data in real time, using the Hadoop cluster, and analysis, visualization and decision making, using online analytical processing and automatic learning techniques.  In addition, a set of guidelines was generated for the implementation of the Big Data architecture designed in financial institutions.  Finally, the Big Data Architecture designed for financial entities was validated based on expert criteria, in which its relevance was demonstrated.