From the Paft to the Fiiture: a Fully Automatic NMT and Word Embeddings Method for OCR Post-Correction
This addresses the time-consuming issue of manual correction for historical texts, offering an unsupervised alternative to rule-based or supervised methods.
The paper tackles the problem of OCR errors in historical corpora by proposing a fully automatic unsupervised method to extract parallel data for training a character-based sequence-to-sequence NMT model, achieving error correction without manual intervention.
A great deal of historical corpora suffer from errors introduced by the OCR (optical character recognition) methods used in the digitization process. Correcting these errors manually is a time-consuming process and a great part of the automatic approaches have been relying on rules or supervised machine learning. We present a fully automatic unsupervised way of extracting parallel data for training a character-based sequence-to-sequence NMT (neural machine translation) model to conduct OCR error correction.