RUS  ENG
Full version
JOURNALS // Sistemy i Sredstva Informatiki [Systems and Means of Informatics] // Archive

Sistemy i Sredstva Inform., 2022 Volume 32, Issue 3, Pages 103–115 (Mi ssi846)

This article is cited in 1 paper

Database of hierarchies of logical-semantic relations: User interface

A. A. Durnovoa, O. Yu. Inkovaba, N. A. Popkovaa

a Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
b University of Geneva, 22 Bd des Philosophes, CH-1205 Geneva 4, Switzerland

Abstract: The paper describes user features of the Database of Hierarchies of Logical-Semantic Relations (SBDH LSR), a new tool for annotating discourse relations, developed at FRC CSC RAS. The article lists annotations stages in the SBDH LSR: selection of the context for annotation, possible modifications, annotation of the LSR marker and of LSR itself. In a single annotation, the formed annotations reflect three important theoretical aspects of discourse analysis: the LSR marker (or its absence); the hierarchy of LSR; and the boundaries of text fragments directly connected by LSR. The database's interface allows one to edit, complete, or filter the annotated information: to save annotations' previous states and, if necessary, go back to one of them, to recover deleted annotations, to use already created annotations to form the new ones. The results obtained with the SBDH LSR have already allowed us to clarify some theoretical points, in particular, those formulated by the Rhetorical Structure Theory. Further analysis of the examples and statistical capabilities of the SBDH LSR will make it possible to establish regularities in the deep LSR hierarchies. This information might be used in cognitive research, both monolingual and contrastive, as well as for automatic text processing and artificial intelligence development.

Keywords: supracorpora database, corpus of text annotations, discourse relations, connector.

Received: 08.08.2022

DOI: 10.14357/08696527220310



© Steklov Math. Inst. of RAS, 2024