A Corpus Reader and Part-of-Speech Taggers for the MULTEXT-EAST corpus based on NLTK and scikit-learn

This repository contains all resources we developed for integrating the MULTEXT-EAST corpus into NLTK. Additionally some POS-Taggers can be found. For an overview over the capabilities of our work have a look at the file multext_nltk.pdf

The raw data we generated in our evaluation can be found in the csv-files in the results directory, whereas in the csvtools directory the scripts for post-processing the csvs are located.

The file mte.py is the actual corpus reader, it is integrated into NLTK in exactly the same version. The MTEDownloader.py is an alternative to the NLTK downloader, which can also be used to retrieve the corpus.