Lexically Aware Semi-Supervised Learning for OCR Post-Correction

Lexically Aware Semi-Supervised Learning for OCR Post-Correction Shruti Rijhwani1, Daisy Rosenblum2, Antonios Anastasopoulos3, Graham Neubig1 1Language Technologies Institute, Carnegie Mellon University, USA 2University of British Columbia, Canada 3Department of Computer Science, George Mason University, USA srijhwan@cs.cmu.edu, daisy.rosenblum@ubc.ca, antonis@gmu.edu, gneubig@cs.cmu.edu Abstract Much of the existing linguistic data in many languages of the world is locked away in non- digitized books and documents. Optical char- acter recognition (OCR) can be used to produce digitized text, and previous work has demon- strated the utility of neural post-correction methods that improve the results of general- purpose OCR systems on recognition of less- well-resourced languages. However, these methods rely on manually curated post- correction data, which are relatively scarce compared to the non-annotated raw images that need to be digitized. In this paper, we present a semi-supervised learning method that makes it possible to utilize these raw images to improve performance, specifically through the use of self-training, a technique where a model is iteratively trained on its own outputs. In addition, to enforce consistency in the recognized vocabulary, we introduce a lexically aware decoding method that aug- ments the neural post-correction model with a count-based language model constructed from the recognized texts, implemented us- ing weighted finite-state automata (WFSA) for efficient and effective decoding. Results on four endangered languages demonstrate the utility of the proposed method, with rela- tive error reductions of 15%–29%, where we find the combination of self-training and lex- ically aware decoding essential for achieving consistent improvements.1 1 Introduction There is a vast amount of textual data available in printed form (Dong and Smith, 2018). In this paper, we address the task of digitizing printed materials that contain text in endangered lan- guages, that is, languages with small populations 1Data and code are available at https://shrutirij .github.io/ocr-el/. of first-language speakers and limited acquisi- tion among younger speakers. Printed texts in endangered languages come from various sources, including linguistic documentation and cultural and educational books. Extracting text data from these documents is valuable for a multitude of reasons. Automatic dig- itization can aid language documentation, preser- vation, and accessibility efforts by archiving the texts and making them searchable for language learners, teachers, and speakers, contributing to essential resources for community-based language revitalization. Further, most endangered languages are under-represented in natural language process- ing technologies, primarily because there is little to no data available for training and evaluation (Joshi et al., 2020). This challenge can be mit- igated by converting printed materials in these languages to a machine-readable format. Optical character recognition (OCR) systems can be used to produce digitized text, and recent work (Rijhwani et al., 2020) has demonstrated that post-correction improves the performance of existing general-purpose OCR systems on endan- gered languages (an example is in Figure 1). Most state-of-the-art OCR post-correction methods use neural sequence-to-sequence models and rely on considerable resources such as a large number of manual transcriptions (Schnober et al., 2016; Rigaud et al., 2019) or substantial textual data to train a language model (Dong and Smith, 2018). To …

Content truncated to 500 words