Distributional models in the task of hypernym discovery

An approach to the solution of the first task of automatically taxonomy construction for the Russian language is described. This task consists in matching unknown input-words with hypernyms from the existing taxonomy. We show that useful results can be attained using pre-trained distribution models without additional training. © Springer Nature Switzerland AG 2020.

Authors
Yadrintsev V. 1, 2 , Ryzhova A.3 , Sochenkov I. 2
Journal
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Language
English
Pages
338-350
Status
Published
DOI
10.1007/978-3-030-59535-7_25
Volume
12412 LNAI
Year
2020
Organizations
  • 1 Peoples Friendship University of Russia (RUDN University), Moscow, Russian Federation
  • 2 Federal Research Center “Computer Science and Control” of the Russian Academy of Sciences, Moscow, Russian Federation
  • 3 Skolkovo Institute of Science and Technology, Moscow, Russian Federation
Keywords
BERT; Distributional models; ELMO; fastText; Hypernym discovery; Rusvectores; RuWordNet
Date of creation
02.11.2020
Date of change
02.11.2020
Short link
https://repository.rudn.ru/en/records/article/record/65462/
Share

Other records

SPERANSKY AND HIS ROLE IN RUSSIAN JURISPRUDENCE

Article
Grebennikov Valery V., Dubrovin Vladimir V., Zhuravlev Valerii A., Kalinovskaya Victoria S., Shiryov Denis A.
Voprosy Istorii. Общество с ограниченной ответственностью "Журнал "Вопросы истории". Vol. 2020. 2020. P. 255-260