A Crowdsourced Twitterbot Workflow for the Digital Humanities

Using crowdsourcing to generate social media posts for a history bot
Research
Author

PM

Published

January 27, 2015

This work comes out of the Open Humanities Hack day in November 2014 at King’s College London organised by the King’s College London Department of Digital Humanities, the Digitised Manuscripts to Europeana (DM2E) project, the Open Knowledge Foundation and the Open Humanities Working Group.

The idea was to build a World War 1 Diaries Twitter Bot to tweet extracts from the diaries of soldiers and others involved in the war, with tweets released exactly 100 years on from the events described.

The result was @ww1witness, initially populated with kind permission from the online diaries of Bernard Brookes

The following image depicts the workflow we developed. An initial spreadsheet of Open sources was constructed by date. The sources are either text extracts (the Bernard Brookes diary was scraped to create these, for instance) or links to the PDF of a scanned diary page. This data could then be imported as tasks to the OKF’s Crowdcrafting website so that tweet extracts could be created by online contributors. The final stage was a Python script to pull out the completed task runs and post them to Buffer for scheduled dissemination through the Twitter account.

twitterbot workflow

Thanks to Francesca Benatti of the OU for her help on the project.

Script to extract from Crowdcrafting.org and post to Buffer