YAGO

What it is

YAGO is a large semantic knowledge base jointly developed by the Max Planck Institute for Informatics and Télécom ParisTech. The fetched project page says it derives knowledge from Wikipedia, WordNet, and GeoNames and reports more than 10 million entities and 120 million facts.

Key concepts

The page highlights a WordNet-derived taxonomy enriched by Wikipedia categories, more than 350,000 classes, confidence values on relations, temporal and spatial annotations, thematic domains, and extraction from ten language editions of Wikipedia. It reports a manually evaluated confirmed accuracy of 95%.

How you’d use it

Download, browse, or query its cross-domain entity-and-fact graph, including time- and space-aware knowledge, or obtain the source code for further work.

LLM angle

none stated

Pitfalls & lessons

The fetched page explicitly says it is the old YAGO project page and is no longer maintained, directing readers to a newer site. Its scale and accuracy figures should therefore be read as claims about the version represented by this archived page, not necessarily the current project.

Verdict

A broad multilingual knowledge-base resource with taxonomy, confidence, and temporal/spatial context, but this particular source is primarily useful as a historical project overview.

Sources consulted

• https://www.mpi-inf.mpg.de/departments/databases-and-information-systems/research/yago-naga/yago/