Saga: Continuous Construction and Serving of Large Scale Knowledge Graphs

Explore a comprehensive talk on Saga, an end-to-end platform for incremental and continuous construction of large scale knowledge graphs developed at Apple. Delve into the complexities of building such a platform in industrial settings, focusing on strong consistency, latency, and coverage requirements. Learn about the challenges of creating source adapters for ingesting heterogeneous data sources, developing entity linking and fusion pipelines for constructing coherent knowledge graphs with a common controlled vocabulary, updating knowledge graphs with real-time streams, and exposing the constructed knowledge through various services. Discover how Saga powers multiple downstream applications through graph services, including low-latency query answering, graph analytics, ML-biased entity disambiguation, semantic annotation, and graph-embedding services. Gain insights into Saga's large-scale production use for powering various user-facing knowledge features, presented by Ihab Ilyas, Professor in the Cheriton School of Computer Science and NSERC-Thomson Reuters Research Chair on Data quality at the University of Waterloo.