Δ Freebase, Hummingbird’s Left Brain Becomes Wikidata

You may have noticed more and more answer boxes in Google lately. Google’s Hummingbird algorithm is its answer to semantic search — to provide search results that are conversational.  Freebase was first publicly available in March 2007 as a place for people to enter information and facts about the world that can be freely accessed by people and machines. In July 2010, Freebase was purchased by Google to help power the Google Knowledge Graph. It has an API, and Facebook, like many sites, uses Freebase to automatically generate pages such as this Dell page on Facebook.

According to Google+ page for Freebase, the website and Freebase API  will be retired by June 30. So Facebook and similar sites will need to pull their data from somewhere else. The most likely place where websites will turn is Wikidata. Google is transferring its data from Freebase to Wikidata and will offer a new API to take over what is called “entity search” powered by Google’s Knowledge Graph.

But these dates are behind schedule according to the Wikidata calendar for Freebase. Primary sources tools were released April 1. But a new Google Knowledge Graph based API will not be launched until Q2 2015, leaving the Freebase website running until Q3. During this time, Google will work to release data from Freebase to the Wikidata community. Denny Vrandečić, previously the Wikidata founder and director is working for Google on the migration.

According to Bob Schmidt, the details of how Wikidata can be use for SEO and Knowledge Graph are buried in these areas.

Found it! After wading through 29 billion pages of the total CF of hell that is Wikipedia documentation on itself (Wikidata being no exception), here is the link that SEOs, web programmers and DBAs will want to review to understand Wikidata classes, properties and datatypes — in other words, the data schema for the database of “facts” about the world’s people, places and things and apparently the new home of the motherlode fueling Google’s and Yahoo’s knowledge graphs (dare we call it the DMOZ of data?):
https://tools.wmflabs.org/wikidata-exports/miga/

For those not up to the technical challenges of learning the code for data schema to update the database directly, enlisting the help of a Wikipedia editor may be more feasible.