A Proposed Approach for Provenance Data Gathering

Abstract

Data provenance focuses on the source of the data and on the identification of data sources and their transformations undergone over time. This paper proposes a generic method for collecting provenance data, and is a follow-up of a study carried out by the same authors in a Brazilian hemotherapy center. This method is based on the W3C’s Provenance Data Model (PROV-DM), and proposes a way to capture, store and analyze anemia-index provenance data by applying a scientific workflow, together with the management of provenance of knowledge. This is an exploratory, practical and deductive study carried out with real data from 197,551 candidates for blood donors, extracted from reports ranging from 2000 to 2018 provided by a Brazilian hemotherapy center. People identified with high anemia rates were quantified and tagged as not-suitable for blood donations. The inadequate candidates were quantified with the highest rate of anemia, and out of 1011 male candidates and 4039 female candidates, women had the highest levels of inadequate blood donations. At the end of this study, it can be concluded that the generic method for collecting data provenance proposed here can be applied in several areas of knowledge.