Individual Submission Summary

Direct link:

Discovering Traces of an Archived Component of Google’s Knowledge Graph from the Freebase Data Dumps

Thu, September 5, 1:00 to 2:30pm, Sheraton New Orleans Hotel, Floor: Eight, Endymion


The Freebase knowledge base was a major Semantic Web and linked data technology that was acquired by Google in 2010 to support the Google Knowledge Graph, the back-end for the artificial intelligence (AI) services that provide structured answers to queries instead of links to external web resources. After Freebase was shutdown in 2016 following an “incredible journey”, the remnants of Freebase are contained in a data dump of over a billion RDF triples. This paper reconstructs the Freebase archive and makes connections to the parent company by finding “fingerprints” of Freebase throughout various Google products and services. The analysis was conducted in three stages. First, a data mining technique was applied to process the Freebase RDF triples. Next, the Freebase ontology of domains, types, and properties was examined qualitatively. The analysis of the Freebase ontology was used to guide a search for traces of Freebase data in currently active products and services. Although Freebase is no longer publicly maintained, this paper finds Freebase still in use throughout various “knowledge-based” products and services, such as the Cloud Vision application programming interface (API), Google Trends, YouTube, and Arts and Culture. This is followed by a critical discussion of the knowledge-based uses of the Freebase knowledge base and the Google Knowledge Graph, as well as a discussion of the implications for Google’s mission to “organize the world’s information and make it universally accessible and useful”.