MORPHOLOGICAL ANALYZER MYSTEM 3.0

The large part of the Russian National Corpus has automatic morphological markup. It is based on the morphological analyzer Mystem developed in Yandex with some postprocessing of the results (for example, all indeclinable nouns acquire the tag '0', verbs are divided into separate paradigms by aspect, etc.). Recently a new (third) version of Mystem has been released (see https://tech.yandex.ru/mystem/). In this article we give an overview of its capabilities.