DocumentCloud forms — documents published by DocumentCloud's users

Source: the DocumentCloud collection on the Internet Archive (archive.org).
Suite: forms_documentcloud.

DocumentCloud is where journalists and requesters publish the documents they report on or obtain,
many through the Freedom of Information Act and state public-records laws. The items carry no rights
statement, and no one has granted a license for them. Their status depends on who made each one:

- US federal forms, in the public domain (17 U.S.C. § 105), with what their filers entered: FERC
  Form 1 pages filed by DTE Electric, Ameren Missouri (Union Electric) and Ameren Illinois (financial
  figures), and a Chicago Police Department application on the Justice Department's COPS portal;
- state and local government documents, from Pennsylvania, Texas and a city police department's
  grant application: public records, though a state or city may hold copyright in its own works;
- forms made by private organizations: the National Association of Broadcasters' political
  broadcast agreement, the NFL Player Disability & Neurocognitive Benefit Plan's application, and
  the Australian Broadcasting Corporation's Trailblazers application; and
- a mock IRS-style form ("Form OBMA"), whose maker isn't recorded.

The documents were published on DocumentCloud by the newsrooms and groups whose reporting obtained
them, credited here:

- MuckRock: the Pennsylvania Right-to-Know requests and the Chicago Police Department's COPS
  application;
- Campaign for Accountability (Julia Long): the Texas Health and Human Services forms and emails;
- Energy and Policy Institute (Matt Kasper): the FERC Form 1 pages;
- ProPublica (Jeremy Merrill): the NAB political broadcast agreement;
- The Marshall Project (Manuel Villa): the police department's grant application;
- Advocacy for Fairness in Sports (Sheilla Dingus): the NFL players' plan application;
- the Australian Broadcasting Corporation, ABC News: its own Trailblazers application; and
- the Washington Examiner (Emily Babay): the mock "Form OBMA".

Datalab, a company, publishes this benchmark. We use these pages to test how documents are parsed,
not for what they say, and most are blank or factual; we believe this is fair use. Some pages show
the names and work contact details of officials, staff and requesters, as they were published on
DocumentCloud. If you made one of these documents, or appear in one, and want it removed, open an
issue at https://github.com/datalab-to/omni_parse_bench/issues.
