How to Describe Implicit Discourse Relations: The Experience of Creating a Dataset in Russian

Авторы: Гончаров Александр Анатольевич, Ярошенко Полина Владимировна
Журнал: Вестник Московского университета. Серия 9: Филология
Номер: 6
Год издания: 2025
Издательство: Изд-во Моск. ун-та
Местоположение издательства: М.
Первая страница: 24
Последняя страница: 35
DOI: 10.55959/MSU0130-0075-9-2025-48-06-2
Аннотация: The article presents an approach to the description of implicit discourse relations (DRs) and the experience of its application to the creation of a corresponding dataset in Russian. The material consists of texts from two parallel subcorpora of the Russian National Corpus: English and French. Using the method of search with exclusion in parallel texts, a list of translation correspondences (pairs of text spans where the text in one language contains markers of the DR in question, and the text span in Russian, on the contrary, does not contain markers of the DR) was composed. The study was conducted using the example of the DRs of cause and contrast, for which more than 1000 annotated translation correspondences were formed. The groups of features used to describe implicit DRs are examined. The features are illustrated with examples from the dataset.
Добавил в систему: Гончаров Александр Анатольевич

	ИСТИНА	Войти в систему Регистрация
	ИСТИНА ПсковГУ
	Главная Поиск Статистика О проекте Помощь

ИСТИНА