• Factchecks
  • Blog
  • Documentation
  • Browser
  • Home
  • Data in dataCommons knowledge graph

    Large datasets are behind many of today’s machine learning ‘wins’. Through dataCommons open Knowledge Graph, developers will have access to “Data as a service”. NOT downloadable datasets. Answers to queries will include the provenance (source of the data). Choice of which data to use, based on the source, is in developer's control. At this time all the data in the Knowledge Graph is free and unrestricted to use. However in the future, just like the Web, we expect much of the data to be free, though some data might be private to the developer and some may have a cost associated with it. The base data in the dataCommons knowledge graph, including taxonomy is derived from schema.org. dataCommons Knowledge Graph (DCKB) contains entities corresponding to many kinds of places, including countries, cities and other administrative areas.

    Geographical Data

    Geographical data in dataCommons include the list of continents, countries, states, administrative areas, cities, their geocodes (lat/long), area, timezones, elevation. This data is obtained from Wikidata. It is not an attempt to accurately reflect geopolitical situation on the ground or territorial control.

    Historical Weather Data

    Historical Weather Data is based on publicly available weather observation data from NOAA. Weather Observation data includes GHCN daily data. The weather data is presented at a city level based on observations from nearest weather station. All data is made available in metric units.

    Weather Observations

    Following weather measurements are available


    Data about People / Population (USA)

    The data about people is available only for USA in the initial release. It includes the following kinds of data from different sources

    Demographic Data

    Demographic Data from US Census is a subset of data from the United States Census Bureau and contains estimates of data from American Community Survey. The American Community Survey is an annual, nationwide survey of more than 3.5 million households in the U.S. The survey produces statistics on demographic, social, economic, and other characteristics about US population and housing. The data includes ACS 5-year data.This data covers a range of topics of interest and includes information about the following population parameters

    Crime Data

    The crime statistics data is obtained from the FBI (Federal Bureau of Investigation) and is based on Uniform Crime Reporting Program of the FBI. The data contains Crime Statistics from years 2008 to 2017 at City and State level as a social indicator.


    According to the FBI, criminal activities are classified into Violent Crimes: and Property Crimes. They are further classified as below and are available in dataCommons.

    In the FBI’s Uniform Crime Reporting (UCR) Program, violent crime is composed of four offenses: murder and nonnegligent manslaughter, rape, robbery, and aggravated assault. Violent crimes are defined in the UCR Program as those offenses that involve force or threat of force.

    In the FBI’s Uniform Crime Reporting (UCR) Program, Property crime includes the offenses of burglary, larceny-theft, motor vehicle theft, and arson. The object of the theft-type offenses is the taking of money or property, but there is no force or threat of force against the victims.

    In UCR Statistics, the UCR Program cautions against drawing simplistic conclusions as one area is safer than another or one agency is more or less efficient than another based solely on these crime counts.


    Voter Registration Statistics

    Voter registration statistics are available for different US States at county level and is obtained from Secretaries of States / Election Boards for select states from the websites of individual state boards of elections that includes voter registration statistics and political affiliations.

    The regulations regarding voter registration and publication of this data, its format and frequency vary considerably by states. In many states Secretary of State or state election board is required by law to produce a Report of Registration detailing voter registration statistics and party affiliation, periodically on their websites. When those publications are available, such statistics are included in the knowledge graph.

    Voter Statistics Observations

    Voter statistics data include the following observations or metrics.