A Tool for Facilitating OCR Postediting in Historical Documents

TL;DR

This paper introduces a post-editing tool for OCR outputs of historical documents, leveraging language models to correct common errors, thereby improving accuracy while maintaining transparency for human review.

Contribution

The paper presents a novel OCR post-editing tool that suggests corrections based on language model scores, specifically designed for historical documents with unique challenges.

Findings

Successfully corrects common OCR errors in historical texts

Enhances accuracy of OCR outputs with language model suggestions

Maintains transparency and allows human intervention

Abstract

Optical character recognition (OCR) for historical documents is a complex procedure subject to a unique set of material issues, including inconsistencies in typefaces and low quality scanning. Consequently, even the most sophisticated OCR engines produce errors. This paper reports on a tool built for postediting the output of Tesseract, more specifically for correcting common errors in digitized historical documents. The proposed tool suggests alternatives for word forms not found in a specified vocabulary. The assumed error is replaced by a presumably correct alternative in the post-edition based on the scores of a Language Model (LM). The tool is tested on a chapter of the book An Essay Towards Regulating the Trade and Employing the Poor of this Kingdom (Cary ,1719). As demonstrated below, the tool is successful in correcting a number of common errors. If sometimes unreliable, it is also transparent and subject to human intervention.