OCR Error Post-Correction with LLMs in Historical Documents: No Free Lunches
: Kanerva, Jenna; Ledins, Cassadra; Käpyaho, Siiri; Ginter, Filip
: Holdt, Špela Arhar; Ilinykh, Nikolai; Scalvini, Barbara; Bruton, Micaella; Debess, Iben Nyholm; Tudor, Crina Madalina
: Resources and Representations for Under-Resourced Languages and Domains
Publisher: University of Tartu Library, Estonia
: 2025
: Proceedings of the Third Workshop on Resources and Representations for Under-Resourced Languages and Domains (RESOURCEFUL-2025)
: 38
: 47
: 978-9908-53-121-2
: https://aclanthology.org/2025.resourceful-1.8/
: https://research.utu.fi/converis/portal/detail/Publication/506501669
Optical Character Recognition (OCR) systems often introduce errors when transcribing historical documents, leaving room for post-correction to improve text quality. This study evaluates the use of open-weight LLMs for OCR error correction in historical English and Finnish datasets. We explore various strategies, including parameter optimization, quantization, segment length effects, and text continuation methods. Our results demonstrate that while modern LLMs show promise in reducing character error rates (CER) in English, a practically useful performance for Finnish was not reached. Our findings highlight the potential and limitations of LLMs in scaling OCR post-correction for large historical corpora.
:
This work was carried out in the Human Diversity University profilation programme (PROFI-7) of the Research Council of Finland, as well as in the context of several other research projects supported by the Research Council of Finland.
