Context These files were used in the [ALTA 2017 Challenge](http://www.alta.asn.au/events/sharedtask2017/index.html). The competition was hosted by [Kaggle in Class](https://www.kaggle.com/c/alta-2017-challenge). If you use these data, please cite: D. Mollá, S. Cassidy. Overview of the 2017 ALTA Shared Task: Correcting OCR Errors (2017). *Proc. ALTA 2017*. [https://aclanthology.coli.uni-saarland.de/papers/U17-1014/u17-1014](https://aclanthology.coli.uni-saarland.de/papers/U17-1014/u17-1014) These files are available as a Kaggle dataset at [http://kaggle.com/dmollaaliod/](correect-ocr-errors) Content * `convert.py` - python script for converting files * `train_input.csv` - the training set * `train_output.csv` - the training set with all OCR errors corrected * `train_output_bigrams.csv` - the solutions of the training set. This file is the actual output of the conversion script using the file train_output.csv as input. * `test_input.csv` - the test set * `test_baseline_bigrams.csv` - a sample submission file in the correct format. This file contains the set of bigrams found in test_input.csv, and the resulting F1 score is zero (or close to zero according to the evaluation performed by Kaggle in Class). Acknowledgements Special thanks to The National Library of Australia and its [Trove database](http://trove.nla.gov.au/), from where this data set was sourced.