Data witnessing: attending to injustice with data in Amnesty International’s Decoders project

Figures & data

Figure 1. Slide from Amnesty showing types of sources, issues and microtasking operations.

Figure 2. Post-it notes at Decoders workshop.

Figure 3. Summary of Decode Darfur.

Figure 4. Digitised text of an urgent action.

Figure 5. User interface design for Decode Urgent Actions.

Figure 6. Interactive visualisation from Decode Urgent Actions.

Figure 7. Wireframe prototype of training sequence for new volunteers.

Figure 8. Visualisations of settlements identified by machine-learning algorithm (Cornebise et al., 2018).

Figure 9. Seven types of JIV forms identified by Amnesty International.

Figure 10. Graph and table showing quantification of delays from JIV reports.

Figure 11. User interface for Amnesty Decoders’ Troll Patrol project.

Figure 12. Decoders Tweet abuse demo, drawing on machine learning experiments.

Cornebise, J., Worrall, D., Farfour, M., & Marin, M. (2018). Witnessing atrocities: Quantifying villages destruction in Darfur with crowdsourcing and transfer learning. Presented at the AI for Social Good NIPS2018 Workshop, Montréal, Canada.

 Google Scholar