2,775,946,436 tweet IDs of tweets about the COVID-19 pandemic, in 26,914 hourly files from 2020-01-21 to 2023-02-17 (hours in UTC), from the ongoing collection of Emily Chen, Kristina Lerman and Emilio Ferrara (USC Information Sciences Institute). The collection began on 28 January 2020 with Twitter's search API gathering the preceding seven days and then followed accounts and keywords through the streaming API; the lists of both are included. Only IDs are released, no tweet text; the tweets have to be hydrated from Twitter.
295,849,036 tweet IDs of tweets about the 2020 United States presidential election, collected during 2020 and 2021 according to the Dataverse record through the Twitter API with Social Feed Manager by Laura Wrubel and Daniel Kerchner of George Washington University Libraries, in 11 collections: Democratic candidates filter (179,765,251), Democratic candidates user timelines (287,722), Democratic Convention (3,552,268), Election 2020 (17,348,084), Election Day (8,410,300), First Presidential Debate (9/29/2020) (2,712,517), Republican candidates filter (69,814,859), Republican candidates user timelines (40,925), Republican Convention Filter (3,189,455), Second Presidential Debate (10/22/2020) (4,833,216), Vice Presidential Debate (10/7/2020) (5,894,439). The files hold IDs only, no tweet text; the tweets have to be hydrated from Twitter.
283,244,653 tweet IDs of tweets about the 2016 United States presidential election, collected between July 13, 2016 and November 10, 2016 through the Twitter API with Social Feed Manager by Justin Littman, Laura Wrubel and Daniel Kerchner of George Washington University Libraries, in 12 collections: Democratic candidates user timelines (38,006), Democratic Convention (8,343,632), Democratic Party twitter timelines (21,426), Election Day (3,630,769), Candidates and key election hashtags twitter filter (251,376,589), First presidential debate (3,183,202), Republican candidates twitter timelines (98,264), GOP Convention (6,255,127), Republican Party twitter timelines (21,871), Second presidential debate (4,116,525), Third presidential debate (2,984,714), Vice Presidential Debate (3,174,528). The files hold IDs only, no tweet text; the tweets have to be hydrated from Twitter.
Tweet IDs of 7,665,497 tweets about the events in Charlottesville, Virginia in August 2017, collected with Social Feed Manager by Justin Littman (George Washington University Libraries) in two collections: a filter stream on the tracked terms #charlottesville, #standwithcharlottesville, #defendCville, #HeatherHeyer and #UnityCville (3,015,437 IDs) and a Twitter search for the same terms (4,650,060 IDs). Only IDs are released, no tweet text; the tweets have to be hydrated from Twitter. Each collection comes with a Social Feed Manager README.
Tweet IDs of 2,041,399 tweets from the official Twitter timelines of members of the 115th U.S. Congress, collected weekly between 27 January 2017 and 2 January 2019 with Social Feed Manager by Justin Littman (George Washington University Libraries): 1,522,189 IDs from 435 accounts of representatives and 519,210 IDs from 102 accounts of senators, each with the list of accounts collected and a Social Feed Manager README. Only IDs are released, no tweet text; the tweets have to be hydrated from Twitter.
Tweet IDs from two Spanish-language Twitter collections made for a bachelor's thesis on hate speech against the LGBTIQ+ community before and after Elon Musk's takeover of Twitter, for 28 June of each year from 2015 to 2025 (Pride Day): Main (694,606 IDs of tweets related to the LGBTIQ+ community, collected with keywords) and Benchmark (657,117 IDs of random tweets, a 6-minute sample of every hour of those days). Only IDs are released, no tweet text and no labels.
Tweet IDs of 957,956 tweets posted between 1 July and 30 September 2020 that match #metoo and related hashtags (#timesup, #metoostem, #sciencetoo, #metoophd and others), a subset of the #metoo Digital Media Collection of the Schlesinger Library, Radcliffe Institute, Harvard University. The tweets were collected weekly through Social Feed Manager with the Twitter streaming filter. Only IDs are released, no tweet text; the tweets have to be hydrated from Twitter.
Tweet IDs of tweets containing #BlackLivesMatter (768,581 IDs) or #AllLivesMatter (101,575 IDs), collected from the Twitter Gardenhose, a 10% sample of all tweets, between 8 August 2014 and 31 August 2015, deposited by Ryan Gallagher. Only IDs are released, no tweet text; the tweets have to be hydrated from Twitter (the Dataverse record points to the Documenting the Now Hydrator).
690,281 tweet IDs of tweets sent between 31 December 2011 and 31 December 2022 by the Twitter accounts of US state public health agencies, state-level COVID-19 response efforts and CDC- and FDA-affiliated accounts, collected by Samuel R. Mendez (Harvard T.H. Chan School of Public Health) through the Twitter Academic Research API on 9 April 2023. Each ID carries a label for the area served by the account's agency (50 state or territory codes, plus CDC with 59,801 IDs and FDA with 31,040). A list of the 65 accounts with their agency, area and first and last tweet dates in the data accompanies the IDs. The files hold IDs only, no tweet text.
Tweet IDs of 254,363 tweets about the 2022 monkeypox outbreak, posted on Twitter from 7 May 2022 to 23 July 2022, collected and shared by Nirmalya Thakur (version 3 of the Zenodo record 6898178). The IDs come in six files in the order of the tweets' dates: part1 (13,926), part2 (17,705), part3 (17,585), part4 (19,718), part5 (46,718), part6 (138,711). Only IDs are released, no tweet text; the tweets have to be hydrated from Twitter.
Twitter (X)