Optimizing the neural network training for OCR error correction of historical Hebrew texts

Suissa, Omri; Elmalech, Avshalom; Zhitomirsky-Geffet, Maayan

Permalink

https://hdl.handle.net/2142/106546 Copy

Description

Title

Optimizing the neural network training for OCR error correction of historical Hebrew texts

Author(s)

• Suissa, Omri
• Elmalech, Avshalom
• Zhitomirsky-Geffet, Maayan

Issue Date

2020-03-23

Keyword(s)

• Optical character recognition (OCR)
• OCR post-correction
• Neural networks
• Hebrew historical newspapers
• Digital humanities

Abstract

Over the past few decades, large archives of paper-based documents such as books and newspapers have been digitized using Optical Character Recognition. This technology is error-prone, especially for historical documents. To correct OCR errors, post-processing algorithms have been proposed based on natural language analysis and machine learning techniques such as neural networks. Neural network's disadvantage is the vast amount of manually labeled data required for training, which is often unavailable. This paper proposes an innovative method for training a light-weight neural network for Hebrew OCR post-correction using significantly less manually created data. The main research goal is to develop a method for automatically generating language and task-specific training data to improve the neural network results for OCR post-correction, and to investigate which type of dataset is the most effective for OCR post-correction of historical documents. To this end, a series of experiments using several datasets was conducted. The evaluation corpus was based on Hebrew newspapers from the JPress project. An analysis of historical OCRed newspapers was done to learn common language and corpus-specific OCR errors. We found that training the network using the proposed method is more effective than using randomly generated errors. The results also show that the performance of the neural net-work for OCR post-correction strongly depends on the genre and area of the training data. Moreover, neural networks that were trained with the proposed method outperform other state-of-the-art neural networks for OCR post-correction and complex spellcheckers. These results may have practical implications for many digital humanities projects.

Publisher

iSchools

Series/Report Name or Number

iConference 2020 Proceedings

Type of Resource

• text
• image

Language

eng

Permalink

http://hdl.handle.net/2142/106546

Copyright and License Information

Copyright 2020 Omri Suissa, Avshalom Elmalech, and Maayan Zhitomirsky-Geffet

Owning Collections