Building NLP Pipeline for Russian with a Handful of Linguistic Knowledge

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F16%3A10372129" target="_blank" >RIV/00216208:11320/16:10372129 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Building NLP Pipeline for Russian with a Handful of Linguistic Knowledge

Popis výsledku v původním jazyce

This work addresses the issue of building a free NLP pipeline for processing Russian texts from plain text to morphologically and syntactically annotated structures in CONLL format. The pipeline is written in python3. Segmentation is provided by our own module. Mystem with numerous postprocessing fixes is used for lemmatization and morphology tagging. Finally, syntactical annotation is obtained with MaltParser utilizing our own model trained on SynTagRus, which was converted into CONLL format for this purpose, with its morphological tagset being converted into Mystem/Russian National Corpus tagset

Název v anglickém jazyce

Building NLP Pipeline for Russian with a Handful of Linguistic Knowledge

Popis výsledku anglicky

This work addresses the issue of building a free NLP pipeline for processing Russian texts from plain text to morphologically and syntactically annotated structures in CONLL format. The pipeline is written in python3. Segmentation is provided by our own module. Mystem with numerous postprocessing fixes is used for lemmatization and morphology tagging. Finally, syntactical annotation is obtained with MaltParser utilizing our own model trained on SynTagRus, which was converted into CONLL format for this purpose, with its morphological tagset being converted into Mystem/Russian National Corpus tagset

Druh

O - Ostatní výsledky

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2016

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů