Historikertage auf Twitter (2012-2018). Datenreport und Datenset

This data report contains the annotated figures, statistics and visualisations of the project "Die twitternde Zunft. Historikertage auf Twitter (2012-2018)" by Mareike König and Paul Ramisch. In addition, the methodological approach to corpus creation, data cleaning, coding, network and text analysis as well as the legal and ethical considerations of the project are described.

The datasheets contain the dehydrated and annotated tweet ids that were used for our study. With the Twitter API this can be used to hydrate and restore the whole corpus, apart from deleted tweets. There are two versions of the CSV file, one with clean id values, the other where the id values are prepended with an “x”. This prevents certain tools from using scientific notation for the ids and breaking them, with the R library rtweet function read_twitter_csv() this is automatically resolved on import.

The files contain the following data:

       status_id: The Twitter status id of the tweet

       corpus_user_id: A corpus specific id for each user within the corpus (not the Twitter user id)

       hauptkategorie_1: Primary category

       hauptkategorie_2: Primary category 2

       Gender: Gender of the user

       Nebenkategorie: Secondary category

Furthermore, the following boolean variables describe what sub corpus each tweet is in, the main corpus per year that contains of both data sources (TAGS and API) and the yearly sub corpora divided by their data source (TAGS: orig_, API: api_):

You can find the code on R on GitHub: https://github.com/dhiparis/historikertag-twitter.