A deep neural network-based data-driven model for evaluating the recognition of ADR mentions in the texts of the PsyTAR corpus

In this paper, for the first time, the accuracy of solving the problem of recognizing named entities of the ADR type for the PsyTAR corpus was evaluated. Initially, the corpus contained markup for solving the problems of detecting entities, classifying their presence, and linking entities. We carrie...

Full description

Saved in:
Bibliographic Details
Main Authors: Sboev, Alexander, Rylkov, Gleb, Selivanov, Anton, Moloshnikov, Ivan, Rybka, Roman
Format: Conference Proceeding
Language:English
Subjects:
Artificial neural networks
Recognition
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, for the first time, the accuracy of solving the problem of recognizing named entities of the ADR type for the PsyTAR corpus was evaluated. Initially, the corpus contained markup for solving the problems of detecting entities, classifying their presence, and linking entities. We carried out additional marking of the corpus. A neural network based on the XLM-RoBERTa architecture was used as a model for entity recognition. The results of this model are also presented in comparison with the published results for this corpus on other tasks.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0162396