QueDI: From Knowledge Graph Querying to Data Visualization

Abstract

While Open Data (OD) publishers are spur in providing data as Linked Open Data (LOD) to boost innovation and knowledge creation, the complexity of RDF querying languages, such as SPARQL, threatens their exploitation. We aim to help lay users (by focusing on experts in table manipulation, such as OD experts) in querying and exploiting LOD by taking advantage of our target users’ expertise in table manipulation and chart creation. We propose QueDI (Query Data of Interest), a question-answering and visualization tool that implements a scaffold transitional approach to 1) query LOD without being aware of SPARQL and representing results by data tables; 2) once reached our target user comfort zone, users can manipulate and 3) visually represent data by exportable and dynamic visualizations. The main novelty of our approach is the split of the querying phase in SPARQL query building and data table manipulation. In this article, we present the QueDI operating mechanism, its interface supported by a guided use-case over DBpedia, and the evaluation of its accuracy and usability level.