Library Guides: Twitter Research Guide: Twitter Data

Twitter Data

On January 26, 2021, Twitter announced the "Academic Research product track" that allows academic researchers to access historical tweets (as well as other features; see the link for the full details).

Below are links to a few tweet ID datasets that are freely available online, though some strategic Internet searching may yield other options. In order to acquire tweets from the Twitter API based on these saved tweet IDs, you'll need to make use of a tweet hydrator, such as Documenting the Now's Hydrator.

Learn more about Twitter data structure and preservation with the Data Curation Network's Twitter Data Primer.

Pennsylvania Perspectives on the 2016 U.S. Election Tweet IDs
"This dataset of Twitter IDs is part of the Pennsylvania Perspectives on the 2016 Election web archive, a collection of websites and social media documenting the diversity of Pennsylvania perspectives around the 2016 U.S. election. The collection includes dedicated election sections on Pennsylvania news outlet websites, campaign websites and social media communications of Pennsylvania candidates for state and federal political offices, websites and social media from statewide political advocacy groups, and the perspectives of university student organizations found on Facebook. This dataset contains only the IDs of the 17,899,637 tweets collected. Each text file represents the tweets collected from a specific Twitter account or by hashtag" (from the collection description).
Documenting the Now Tweet Datasets
"This site is a catalog of datasets that are publicly available on the web. If you would like to turn these tweet identifier datasets back into the original JSON first download the dataset and then use the Hydrator desktop application, or Twarc if you are comfortable working at the command line" (from the Documenting the Now website description).