dataset of James Joyce's collection of short stories "Dubliners". prepared for the nlp course in hse (https://github.com/vifirsanova/compling).
contains the text tokenized by sentences, pos-tagged sentences with nltk, and results of analyzing the text with spacy library (pos-tagged, named entities, dependencies).
usable for different nlp tasks.