Abstract:

Like any other area of ​​knowledge, education systems produce large volumes of data in a continuous way, especially at the higher education level. Anyone that is interested for extracting useful knowledge from these data, imposes the use of sophisticated processes to do it as a way to associated reduce costs and time. Recently, several technologies have been developed to facilitate the analysis and extraction of knowledge from large databases. Some of them are turning the traditional Web into a big publishing and data consumption platform. Among the paradigms on which the Web has been developing, Semantic Web and Linked Open Data (LOD) have caught our attention as the most adequate fields for discussing strategies to facilitate the production and consumption of LOD on the results of higher education in Brazil. In this context, the literature review has demonstrated that the technologies required for the production and consumption of LOD are already increasingly mature. It was verified that the institutions of higher education produce and publish large amounts of data in the Web with a great social interest. However, although the amount of data has increased, it has been found that most of the available data sets do not meet most of the good practices used on LOD publication.

In this case, the main problems identified were:
1) adoption of proprietary standards;
2) ambiguous terminology; and,
3) absence of metadata.

The literature also showed that the increasing presence of higher education data on the Web brings new discussions and new actors to the evaluation of higher education. Some related work shows that producing open and connected data goes beyond the mere technical issue of assigning a uniform resource identifier (URI) and a syntax description to a dataset. On the contrary, they highlight the importance of adding explicit semantics and information about the context in which such data sets were produced. In this work we discuss the importance of pragmatics for publishing and consuming LOD on the evaluation of higher education in Brazil. Additionally, we analyze the most important Semantic Web technologies related to the linked data, highlighting the use of ontologies as formal conceptualizations of reference that are comprehensible for both humans and computers. Finally, we present a core ontology for the National Higher Education Assessment System in Brazil and we demonstrate its use to provide a reference conceptualization for the publication and consumption of LOD within the domain.

Keywords:

Education, Semantic Web, Linked Open Data, Ontologies Based Systems, Brazil Higher Education.