Visual SPARQL querying based on extended filter/flow graphs

Abstract

SPARQL is currently the major query language for the Semantic Web. However, writing SPARQL queries is not an easy task and requires some understanding of technologies like RDF. In order to enable users without this knowledge to query linked data, visual interfaces are required that hide the SPARQL syntax and provide graphical support for query building. Based on the concept of extended filter/flow graphs, we present a novel approach for visual querying that addresses the unique specifics of SPARQL and RDF. In particular, it enables the creation of SELECT and ASK queries, though it can also be used for other query forms. In contrast to related work, the users do not need to provide any structured text input but can create the queries entirely with graphical elements. Our approach supports most features of SPARQL and hence also the construction of complex query expressions. It has been implemented in a visual querying framework and tested on different RDF datasets, including DBpedia that is used as an example in this paper. Since the filter/flow concept is empirically well-founded, we expect our approach to be very usable, which is additionally supported by the results of a qualitative user study we conducted.