Context: Portuguese is a romance language that is the native language of over 215 million speakers worldwide. Like Spanish, English and French, it was the language of both its country of origin and also that country’s colonial possessions. This corpus contains examples of historical Portuguese written between 1500 and 1936, both in Portugal and Brazil. Content: The corpus contains complete Portuguese manuscripts published from 1500 to 1936 divided into 5 sub-corpora per century (summarized in the table below). The part of speech (POS) of words in this corpus was tagged using [TreeTagger](http://www.ims.uni-stuttgart.de/forschung/ressourcen/werkzeuge/treetagger.html). You can find more information on this corpus on the [Colonia homepage](http://corporavm.uni-koeln.de/colonia/index.html). Century Texts Tokens 16th 13 399,245 17th 18 709,646 18th 14 425,624 19th 38 2,490,771 20th 17 1,132,696 Total 100 5,157,982 Texts are balanced in terms of the variety, consisting of 48 European Portuguese texts and 52 Brazilian Portuguese texts. You can find more information in the paper that describes the corpus. The complete inventory of texts is here and more detail regarding annotation can be found here. Part of Speech (POS) Tags The works in this corpus have been automatically tagged for their part of speech (POS). The tagset used to annotate the corpus is presented in the table below. It contains not only the classic POS tags (e.g. V, DET, N) but also a couple of compound tags, such as the combination of preposition plus determiner as (PREP+DET) or verb plus pronoun (V+P). The tool used to annotate the corpus was TreeTagger. Category POS Example Adjective ADJ bonita Adverb ADV muita Determiner DET os Cardinal CARD primeiro Noun NOM mesa Pronoun P eles Preposition PRP de Verb V fazer Interjection I Oh! Commas VIRG , Punctuation SENT . Studies report that TreeTagger achieves performance higher than 95% accuracy in attributing the correct POS tag and lemma of a token. Acknowledgements: If you use this corpus in your work, please cite this paper: Zampieri, M. and Becker, M. (2013) Colonia: Corpus of Historical Portuguese. In: ZSM Studien, Special Volume on Non-Standard Data Sources in Corpus-Based Research. Volume 5. Shaker. Inspiration: * What changes have occurred in Portuguese over time? Have words changed? Syntactic structures? How grammatical agreement is expressed? * Can you create a classifier which can classify the era and unseen work is from? * Using the part of speech tags in this tagger, can you train a new tagger and run it over the Brazilian Portuguese Literature Corpus linked below? You may also like: [A 3.7 million word literary corpus of Brazilian Portugese](https://www.kaggle.com/rtatman/brazilian-portuguese-literature-corpus)