The size of knowledge graphs has reached the scale where centralised analytical approaches have become infeasible. Recent technological progress has enabled powerful distributed in-memory analytics that have been shown to work well on simple data structures. However, the application of such distributed analytics approaches on semantic knowledge graphs lags significantly behind. To advance both scalability and accuracy of large-scale knowledge graph analytics to a new level, foundational research on methods leveraging distributed in-memory computing and semantic technologies in combination with advancements in analytics approaches is indispensable.

This Lecture introduces SANSA Scalable Semantic Analytics Stack that provides support for

• efficient data distribution and semantics-aware computation of latent resource embeddings for knowledge graphs;
• adaptive distributed querying;
• efficient self-optimising inference execution plans; and 
• efficient distributed machine learning on semantic knowledge graphs of extremely large scale.