Unifying Faceted Search and Analytics over RDF Knowledge Graphs

The formulation of analytical queries over Knowledge Graphs in RDF is a challenging task that presupposes familiarity with the syntax of the corresponding query languages and the contents of the graph. To alleviate this problem, we introduce a model for aiding users in formulating analytic queries over complex, i.e., not necessarily star schema-based, RDF Knowledge Graphs. To come up with an intuitive interface, we leverage the familiarity of users with Faceted Search systems. In particular, we start from a general model for Faceted Search over RDF data, and we extend it with actions that enable users to formulate analytic queries, too. Thus, the proposed model can be used not only for formulating analytic queries but also for exploratory purposes, i.e., for locating the desired resources in a Faceted Search manner. We describe the model from various perspectives, i.e., (1) we propose a generic user interface for intuitively analyzing RDF Knowledge Graphs, (2) we define formally the state space of the interaction model and the required algorithms for producing the user interface actions, (3) we present an implementation of the model that showcases its feasibility, and (4) we discuss the results of an evaluation with users that provides evidence for the acceptance of the method by users. Apart from being intuitive for end users, another distinctive characteristic of the proposed model is that it allows the gradual formulation of complex analytic queries (including nested ones).