Data Variety Modeling: A Case of Contextual Diversity Identification from a Bottom-up Perspective.

Variety is a property related to data diversity in Big Data Systems (BDS) that comprises several cases, such as structural diversity (variety in data types), source diversity (variety in the way data are produced), etc. Recently, adding contextual information allows more complex analyses, which open the possibility of modeling variety thinking of reuse. In this article, we introduce a proposal for modeling variety by following a dual data processing perspective. We exemplify the case of one of these perspectives by identifying and modeling contextual variations in a particular domain problem.