Master theses

RMLFish - A fish eye view of RML mapping processes

Keywords: Semantic Web, Linked data, RML mapping, UI, Visualization

Supervision: Ruben Verborgh Ben De Meester

Students: max 1

Several solutions exist for the generation of Knowledge Graph (KG) data from heterogeneous data sources such as OBDA, OBDI, and mapping process. One such type, the mapping process, is a specialized research area of KG construction in our IDLab’s KnoWS group. Currently, we have state-of-the-art solutions to generate quality KG from heterogeneous data sources such as RMLMapper and RMLStreamer. These tools use a mapping document, RML, to map heterogeneous data to KG. However, the RML documents are utilized without any optimization to the given mapping processes. Moreover, when multiple RML documents are used to process a dataset, there is no control in place to ensure there is no redundant mapping of the same data records. Hence, it is desirable to have a validator/optimizer for the RML mapping documents.