Methods of Solving the Problem of Coreference and Searching for Noun Phrases in Natural Languages
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F26%3A3UGB7FF4" target="_blank" >RIV/00216208:11320/26:3UGB7FF4 - isvavai.cz</a>
Result on the web
<a href="http://dx.doi.org/10.1134/S1064230725700108" target="_blank" >http://dx.doi.org/10.1134/S1064230725700108</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1134/S1064230725700108" target="_blank" >10.1134/S1064230725700108</a>
Alternative languages
Result language
angličtina
Original language name
Methods of Solving the Problem of Coreference and Searching for Noun Phrases in Natural Languages
Original language description
Abstract: Coreference is a task in the field of natural language processing aimed at linking words and phrases in a text that point to the same extra-linguistic object or referent. It is applicable in text summarization, question answering, information retrieval, and dialog systems. In this paper, the existing methods for solving the coreferencing problem are dissected and a method based on the application of a two-stage machine learning model is proposed. The language model converts text tokens into vector representations. Then, for each pair of tokens, based on their vector representations, an estimate of the probability of finding these tokens either in one noun phrase or in two coreference noun phrases is computed. Thus, the method simultaneously searches for noun phrases and predicts the coreference relationship between them. © Pleiades Publishing, Ltd. 2025.
Czech name
—
Czech description
—
Classification
Type
J<sub>SC</sub> - Article in a specialist periodical, which is included in the SCOPUS database
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
—
Continuities
—
Others
Publication year
2025
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Name of the periodical
Journal of Computer and Systems Sciences International
ISSN
1064-2307
e-ISSN
—
Volume of the periodical
64
Issue of the periodical within the volume
1
Country of publishing house
US - UNITED STATES
Number of pages
15
Pages from-to
121-135
UT code for WoS article
—
EID of the result in the Scopus database
2-s2.0-105004467498