Discovering Implicational Knowledge in Wikidata

Talk by Maximilian Marx

Location: APB 3027
Start: 13. June 2019 at 1:00 pm
End: 13. June 2019 at 1:30 pm
Research group: Knowledge-Based Systems
Event series: KBS Seminar
iCal

Knowledge graphs have recently become the state-of-the-art tool for representing the diverse and complex knowledge of the world. Among the freely available knowledge graphs, Wikidata stands out by being collaboratively edited and curated. Amidst the vast numbers of facts, complex knowledge is just waiting to be discovered, but the sheer size of Wikidata makes this infeasible for human editors. We apply Formal Concept Analysis to efficiently identify and succinctly represent comprehensible implications that are implicitly present in the data. As a first step, we describe a systematic process to extract conceptual knowledge from Wikidata's complex data model, thus providing a method for obtaining large real-world data sets for FCA. We conduct experiments that show the principal feasibility of the approach, yet also illuminate some of the limitations, and give examples of interesting knowledge discovered.

This will be a rehearsal talk for ICFCA-2019 (20 minutes including questions).