More and more companies are building and using knowledge graphs. Graphs are actively using to study research (Amazon, Netflix), to analyse the stock market (Goldman Sachs), search (Yandex, Google), and even to search for new molecules. Often, the creation and populating of the graph with information is automatic or semi-automatic. Subsequently, the final functions of the graph (quality of prediction, validity of derived facts) depend on the quality of data in the graph. So ensuring the graph data quality is a significant process, and it must be approached responsibly.
In this talk, I will present different approaches to the graph data quality control.