Multi-modular domain-tailored OCR post-correction

Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 2716–2726 Copenhagen, Denmark, September 7–11, 2017. c⃝2017 Association for Computational Linguistics Multi-modular domain-tailored OCR post-correction Sarah Schulz and Jonas Kuhn Institute for Natural Language Processing (IMS) University of Stuttgart firstname.lastname@ims.uni-stuttgart.de Abstract One of the main obstacles for many Dig- ital Humanities projects is the low data availability. Texts have to be digitized in an expensive and time consuming pro- cess whereas Optical Character Recogni- tion (OCR) post-correction is one of the time-critical factors. At the example of OCR post-correction, we show the adap- tation of a generic system to solve a spe- cific problem with little data. The system accounts for a diversity of errors encoun- tered in OCRed texts coming from differ- ent time periods in the domain of litera- ture. We show that the combination of dif- ferent approaches, such as e.g. Statisti- cal Machine Translation and spell check- ing, with the help of a ranking mecha- nism tremendously improves over single- handed approaches. Since we consider the accessibility of the resulting tool as a crucial part of Digital Humanities col- laborations, we describe the workflow we suggest for efficient text recognition and subsequent automatic and manual post- correction. 1 Introduction Humanities are no longer just the realm of schol- ars turning pages of thick books. As the worlds of humanists and computer scientists begin to in- tertwine, new methods to revisit known ground emerge and options to widen the scope of research questions are available. Moreover, the nature of language encountered in such research attracts the attention of the NLP community (Kao and Juraf- sky (2015), Milli and Bamman (2016)). Yet, the basic requirement for the successful implemen- tation of such projects often poses a stumbling block: large digital corpora comprising the textual material of interest are rare. Archives and individ- ual scholars are in the process of improving this situation by applying Optical Character Recog- nition (OCR) to the physical resources. In the Google Books1 project books are being digitized on a large scale. But even though collections of literary texts like Project Gutenberg2 exist, these collections often lack the texts of interest to a spe- cific question. As an example, we describe the compilation of a corpus of adaptations of Goethe’s Sorrows of the young Werther which allows for the analysis of character networks throughout the pub- lishing history of this work. The success of OCR is highly dependent on the quality of the printed source text. Recognition er- rors, in turn, impact results of computer-aided re- search (Strange et al., 2014). Especially for older books set in hard-to-read fonts and with stained paper the output of OCR systems is not good enough to serve as a basis for Digital Humanities (DH) research. It needs to be post-corrected in a time-consuming and cost-intensive process. We describe how we support and facilitate the manual post-correction process with the help of informed automatic post-correction. To account for the problem of relative data sparsity, we …

Content truncated to 500 words