Um processo para conversão e publicação de dados para modelo rdf seguindo os princípios de Linked Data

Abstract

After the initial period of great enthusiasm through the publication of Databases in linked data, the scientific community decides now evaluate the bases as their quality. However, some problems have often presented, such as: lack of a process which encourages, guides and contributes with the improvement of publication and reuses of data connected to the web and can not identify, verify and/or evaluate the source and workflow conducted prior to publication. Therefore, this study aims to propose a process that encourages, guides and contributes with the improvement of publication and reutilization of data connected in the Web, as well as simplifies and recommends the use of some tools, standards, principles and good practices for transformation, interconnection, exposure, and sharing of data resources in RDF model, considering Linked Data patterns. It was performed an example of a real application involving open data and transparency of the Brazilian government related to the combat of fraudulent companies of bidding public commerce to highlight the contribution of the proposed approach. Besides that, the execution of a workflow ETL (Extraction, Transformation and Load) by Pentaho tool and ETL4LOD plugin to convert data sources of various formats, to RDF, in order to demonstrate that this tool is able to automate the generation process and update of RDF, as well as SPARQL queries.