How can we achieve sufficient levels of transparency and fairness for (humanities) research based on historical newspapers? Which concrete measures should be taken by data providers such as libraries, research projects and individual researchers? We approach these questions from the vantage point that digitised newspapers are complex sources with a high degree of heterogeneity caused by a long chain of processing steps, ranging, e.g., from digitisation policies, copyright restrictions to the evolving performance of tools for their enrichment such as OCR or article segmentation. Overall, we emphasise the need for careful documentation of data processing, research practices and the acknowledgement of support from institutions and collaborators.

doi.org/10.4230/DagRep.12.7.112
Computational approaches to digitised historical newspapers
Human-Centered Data Analytics

Beelen, K., Chambers, S., Düring, M., Hollink, L., Jänicke, S., Jean-Caurant, A., … Pfanzelter, E. (2023). Fairness and transparency throughout a digital humanities workflow: Challenges and recommendations. Dagstuhl Reports, 7(12), 144–174. doi:10.4230/DagRep.12.7.112