INTERNET-CORPORA FOR FINNO-UGRIC LANGUAGES OF RUSSIA

Abstract

Digital language corpora have long become one of the most important tools in linguistic research; a new methodological approach, known as corpus linguistics, has been based on corpora. While comprehensive corpora exist for the major European Uralic languages (Hungarian, Finnish, Estonian), the smaller Uralic languages of Russia did not have comparable resources until recently. In this paper, I present digital corpora recently developed for the largest Uralic languages of Russia: Udmurt, Komi-Zyrian, Meadow Mari, Erzya and Moksha. The corpora comprise digital texts available on the internet, which were collected and processed by the author. Two corpora were created for each language: a social media corpus and a non-social-media (“main”) corpus. Both kinds of texts were automatically morphologically analyzed; the social media texts were additionally filtered and anonymized. I will outline the development process of these corpora, as well as present their features and possible applications. All corpora described in the paper are equipped with a web-based user interface and are publicly available at http://volgakama.web-corpora.net/.