Confusion table based on OCRed newspaper data
Myriam C. Traub, Jacco van Ossenbruggen · Figshare · 2015
The file contains a table with character confusions we found by comparing OCRed newspaper pages from the digital archive of the National Library of The Netherlands to a ground truth data set (http://lab.kbresearch.nl/static/html/impact.html). The creation and use of this table is discussed in the following publication: Traub, M. C. and van Ossenbruggen, J. R. and Hardman, L., Impact Analysis of OCR Quality on Research Tasks in Digital Archives, Proceedings of the International Conference on Theory and Practice of Digital Libraries (TPDL, 2015)