Y-TAGGRU

Y-TAGGRU is a Russian POS tagger that identifies various types of phrases employing 1) a smart machine learning algorithm based on Bayes theorem, 2) matching against dictionary terms, as well as 3) lemmas, 4) endings. The description of the algorithm is given in the paper

Yatsko, V.A. Another tagger. 2018. DOI: 10.13140/RG.2.2.28859.52007/1. URL: https://www.researchgate.net/publication/326404584_Another_tagger

The tagger underlies the functioning of Y-Mining system that has been designed to identify customers' sentiments about phones, films, and hotel services.

On this site the tagger is distributed as a standalone application for demonstration purposes. The application has the following interface.

The user can open a txt. file (UTF-8 encoding) to get the following results

- POS tags assigned to each word,

- a list of tags and methods by means of which they were recognized,

- statistical information about the number of tags, words, and tokens.

To get Y-TAGGRU go to Downloads section.