Understanding the Structure of Knowledge Graphs with ABSTAT Profiles

Tracking #: 2705-3919

This paper is currently under review
Blerina Spahiu
Matteo Palmonari
Renzo Arturo Alva Principe
Anisa Rula

Responsible editor: 
Karl Hammar

Submission type: 
Full Paper
While there has been a trend in the last decades for publishing large-scale and highly-interconnected Knowledge Graphs in both academia and industry, people and organizations get overwhelmed by the daunting task of understanding as a result of their size and complexity. Data profiling and summarization approaches have been proposed to summarize large Knowledge Graphs into concise and meaningful representation, so that they can be more easily visualized, processed, and managed. While many of these approaches have been proposed to mitigate the problem of data understanding, little attention has been provided to quantify their usefulness for data understanding. In this paper, we discuss how ABSTAT, a data profiling tool supports users in understanding big and complex knowledge graphs. We propose a methodology to evaluate how a profiling tool helps users in understanding the data through the assignment of cognitive tasks. In particular, we construct and present a user study based on query completion tasks where users make use of ABSTAT profiles to complete their queries. Furthermore, we demonstrate the impact of minimalization, a distinctive ABSTAT feature, on the conciseness of the schema-level summaries. The experimentation shows that our profiling framework provides informative and concise profiles, helping users understanding Knowledge Graphs effortlessly.
Full PDF Version: 
Under Review