Annotation-Based Method for Linking Local and Global Knowledge Graphs
Patrícia Raia Nogueira Cavoto, André Santanchè · ONTOBRAS · 2015
In the last years, the use of data available in as Linked Open Data and Ontologies are increasing faster and bringing with them the popularization of the graph structure to represent information networks. One challenge, in this context, is how to link local and global knowledge graphs. This paper presents an approach to address this problem through an annotation-based method to link a local graph database to global graphs. Different from related work, the local graph is not derived from a static dataset, but it is a dynamic graph database evolving along the time, containing connections (annotations) with global graphs that must stay consistent during its evolution. We applied this method over a dataset with more than 44,500 nodes, annotating them with the values found in DBpedia and GeoNames. The proposed method is an extension of our ReGraph 1 framework that bridges relational and graph databases, keeping both integrated, synchronized and in their native representations, with minimal impact in the current infrastructure.