This is a demo for OpenNLP POSTagger trained on the Ancient Greek Dependency Treebank v2.0. Documentation about the tagger can be found on the corresponding openNLP website. The accuracy measured is 0.78. The service can also be accessed via a POST request to

where the body of the message is the text to tag.

If you use the command line, you can type:

curl -X POST --data "ἔφυγέ τε καὶ πόλεμον ἐπήγαγε τῇ πατρίδι."

Note that the unit to analyze is always the sentence: each text should contain at least one sentence, i.e., a string of words whose last character is a full stop, a semicolon, or the middle dot.