Formality annotated Japanese-English parallel data

This is a dataset containing Japanese-English parallel data, annotated according to the formality of the Japanese sentences. The formality classification was automatic and was based on a rule-based formality classifier proposed by Feely et al. 19. The base unannotated parallel corpora that we used can be found at http://www.phontron.com/japanese-translation-data.php, a selection of Japanese-English parallel corpora curated by Graham Neubig.