Unsupervised Transcription of Historical Documents

Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics, pages 207–217, Sofia, Bulgaria, August 4-9 2013. c⃝2013 Association for Computational Linguistics Unsupervised Transcription of Historical Documents Taylor Berg-Kirkpatrick Greg Durrett Dan Klein Computer Science Division University of California at Berkeley {tberg,gdurrett,klein}@cs.berkeley.edu Abstract We present a generative probabilistic model, inspired by historical printing pro- cesses, for transcribing images of docu- ments from the printing press era. By jointly modeling the text of the docu- ment and the noisy (but regular) process of rendering glyphs, our unsupervised sys- tem is able to decipher font structure and more accurately transcribe images into text. Overall, our system substantially out- performs state-of-the-art solutions for this task, achieving a 31% relative reduction in word error rate over the leading com- mercial system for historical transcription, and a 47% relative reduction over Tesser- act, Google’s open source OCR system. 1 Introduction Standard techniques for transcribing modern doc- uments do not work well on historical ones. For example, even state-of-the-art OCR systems pro- duce word error rates of over 50% on the docu- ments shown in Figure 1. Unsurprisingly, such er- ror rates are too high for many research projects (Arlitsch and Herbert, 2004; Shoemaker, 2005; Holley, 2010). We present a new, generative model specialized to transcribing printing-press era documents. Our model is inspired by the un- derlying printing processes and is designed to cap- ture the primary sources of variation and noise. One key challenge is that the fonts used in his- torical documents are not standard (Shoemaker, 2005). For example, consider Figure 1a. The fonts are not irregular like handwriting – each occur- rence of a given character type, e.g. a, will use the same underlying glyph. However, the exact glyphs are unknown. Some differences between fonts are minor, reflecting small variations in font design. Others are more severe, like the presence of the archaic long s character before 1804. To address the general problem of unknown fonts, our model (a) (b) (c) Figure 1: Portions of historical documents with (a) unknown font, (b) uneven baseline, and (c) over-inking. learns the font in an unsupervised fashion. Font shape and character segmentation are tightly cou- pled, and so they are modeled jointly. A second challenge with historical data is that the early typesetting process was noisy. Hand- carved blocks were somewhat uneven and often failed to sit evenly on the mechanical baseline. Figure 1b shows an example of the text’s baseline moving up and down, with varying gaps between characters. To deal with these phenomena, our model incorporates random variables that specifi- cally describe variations in vertical offset and hor- izontal spacing. A third challenge is that the actual inking was also noisy. For example, in Figure 1c some charac- ters are thick from over-inking while others are ob- scured by ink bleeds. To be robust to such render- ing irregularities, our model captures both inking levels and pixel-level noise. Because the model is generative, we can also treat areas that are ob- scured …

Content truncated to 500 words