All datasets generated from the experimental crowdsourcing projects hosted on this platform are made available under a CC0 license.

We have created some (beta) versions of Jupyter notebooks that you can run to view and visualise data created with the project. You can adapt these to download additional information.

Following a results analysis process, our results are stored as Web Annotations and are made available both for download on this page and via the API.

For each project, there are three types of dataset available, all of which can be downloaded in JSON or CSV format:

  • Tasks: The initial task input.
  • Task Runs: All contributions made so far.
  • Results: The final results data, following any analysis of the completed tasks.

We are keen for these datasets to be used in innovative ways, perhaps to further research into new technologies. For instance, the digitised playbills alongside the final results might prove useful for testing pattern recognition applications, such as those using OCR or NER technologies.

Head over to the forum, email or contact @LibCrowds to let us and others know how you have made use of the data, or if you have any further enquiries.


Download all task and contribution data.

Results Download
Tags Download