Conceptual Modeling of Big Data Extract Processes with UML

Abstract

Big Data is a popular term used to define the storage and processing of high volumes of data. The main aim is to assist companies to make better business decisions. There is a lot of research about developing systems and techniques to deal with Big Data and, since a picture is worth a thousand words, the authors usually present diagrams of their proposals. There is, in this regard, a lack of a standardized format to model Big Data; thus, this paper intends to promote the use of the Unified Modeling Language (UML) for modeling Big Data scenarios. In this paper, the use of UML for modeling the extract process of Big Data is presented. UML is a standard that provides several useful elements for representing the main ideas during the design of a system. Some systems require certain concepts that are not covered by UML. For these cases, the metamodel of UML can be extended using a mechanism called stereotyping. In this paper, we propose five new stereotypes and the use of three others proposed in a previous research. To provide a better understanding, we have modeled three tools used in the Big Data extract process. The results state a format based on a standard.